The DOI matches its Crossref record; 140 of the 168 references it cites were found in a registry or an index; 10 of the 54 sources its claims cite were checked for support.
Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing
Open accessAbstract
With the breakthroughs in deep learning, the recent years have witnessed a booming of artificial intelligence (AI) applications and services, spanning from personal assistant to recommendation systems to video/audio surveillance. More recently, with the proliferation of mobile computing and Internet of Things (IoT), billions of mobile and IoT devices are connected to the Internet, generating zillions bytes of data at the network edge. Driving by this trend, there is an urgent need to push the AI frontiers to the network edge so as to fully unleash the potential of the edge big data. To meet this demand, edge computing, an emerging paradigm that pushes computing tasks and services from the network core to the network edge, has been widely recognized as a promising solution. The resulted new interdiscipline, edge AI or edge intelligence (EI), is beginning to receive a tremendous amount of interest. However, research on EI is still in its infancy stage, and a dedicated venue for exchanging the recent advances of EI is highly desired by both the computer system and AI communities. To this end, we conduct a comprehensive survey of the recent research efforts on EI. Specifically, we first review the background and motivation for AI running at the network edge. We then provide an overview of the overarching architectures, frameworks, and emerging key technologies for deep learning model toward training/inference at the network edge. Finally, we discuss future research opportunities on EI. We believe that this survey will elicit escalating attentions, stimulate fruitful discussions, and inspire further research ideas on EI.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
10 of 54 cited sources checked · 4 supported · 0 flagged · accuracy not yet measured
Every source the paper cites for this claim was checked and supports it.
In the paper
“Stamoulis et al. [106] cast the adaptive DNN model selection issue as a hyperparameter optimization problem by taking into account the accuracy and communication constraints imposed by the devices.Then, a Bayesian optimization (BO) is adopted to solve this problem, achieving an improvement by up to 6× in terms of minimum energy per image under accuracy constraints.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > Park et al.
What the sources say
- Supportedcites [106] 106. D. Stamoulis et al., "Designing adaptive neural networks for energy-constrained image classification," in Proc. ACM ICCAD, 2018, Art. no. 23. source
Section 4.1 (p. 4) formulates adaptive CNN design, including network selection, as a hyperparameter optimization problem. Section 4.2 (p. 4) says the authors adapt Bayesian optimization to solve it, and the Conclusion (p. 8) reports up to 6× lower minimum energy per image under accuracy constraints than previously published methods. (LLM, conf=0.94)
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes- Supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“The result varies with architectures and transfer techniques.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Not checked — no full text of the sourcecites [62] 62. R. Sharma, S. Biookaghazadeh, B. Li, and M. Zhao, "Are existing knowledge transfer techniques effective for deep learning with edge devices?" in Proc. IEEE Int. Conf. Edge Comput. (EDGE), Jul. 2018, pp. 42-49. source
- Could not verifycites [44] 44. Q. Chen, Z. Zheng, C. Hu, D. Wang, and F. Liu, "Data-driven task allocation for multi-task transfer learning on the edge," in Proc. IEEE 39th Int. Conf. Distrib. Comput. Syst. (ICDCS), 2019. sourceExperimental — accuracy not yet measured
Section 5.2.2 (p. 6) reports accuracy for one architecture under different distributed-training conditions, but does not compare architectures or transfer-learning techniques. Section 6 (p. 6) calls studies of other DNN architectures future work. Searches for transfer learning and variation across architectures or transfer techniques found no passage addressing the claim. (LLM, conf=0.76)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Cachier can increase responsiveness by 3× or more.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) Edge Caching:
What the sources say
- Could not verifycites [93] 93. U. Drolia, K. Guo, J. Tan, R. Gandhi, and P. Narasimhan, "Cachier: Edge-caching for recognition applications," in Proc. IEEE ICDCS, Jun. 2017, pp. 276-286. sourceExperimental — accuracy not yet measured
The Results section discusses Edgent’s inference latency, but does not report a result for Cachier. Searches for “Cachier” and “responsiveness three times” found no passage supporting a 3× improvement. (LLM, conf=0.77)
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes- Could not verify
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Data Generated at the Network Edge Need AI to Fully Unlock Their Potential”
- Also stated in: § A. Motivation and Benefits of Edge Intelligence · boosted
“AI will be functionally necessary due to its ability to quickly analyze those huge data volumes and extract insights from them for high-quality decision making.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > A. Motivation and Benefits of Edge Intelligence
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Also stated in: § A. Motivation and Benefits of Edge Intelligence · boosted
In the paper
“Cisco estimates that nearly 850 ZB will be generated by all people, machines, and things at the network edge by 2021 [3].”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > I. I N T R O D U C T I O N
What the sources say
- Not matchedcites [3] 3. Cisco Global Cloud Index: Forecast and Methodology, 2016-2021, White Paper. [Online]. Available: https://www.cisco.com/c/en/us/ solutions/collateral/service-provider/global- cloud-index-gci/white-paper-c11-738085.html
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes- Not matched
In the paper
“Lin et al. [53] observe that 99.9% of the gradient exchange in distributed SGD (DSGD) are redundant, which demonstrates the power of gradient compression.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 3) Gradient Compression:
What the sources say
- Not checked — no full text of the sourcecites [53] 53. Y. Lin, S. Han, H. Mao, Y. Wang, and W. J. Dally, "Deep gradient compression: Reducing the communication bandwidth for distributed training," 2017, arXiv:1712.01887. [Online]. Available: https://arxiv.org/abs/1712.01887 source
Load-bearingStrength:boosted· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIn the paper
“Specifically, while preserving client's own privacy, the training loss can be reduced by sharing the models among clients, comparing to training solely on their own inputs.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > C. Enabling Technologies
What the sources say
- Not checked — no full text of the sourcecites [46] 46. R. Shokri and V. Shmatikov, "Privacy-preserving deep learning," in Proc. 22nd ACM SIGSAC Conf. Comput. Commun. Secur., 2015, pp. 1310-1321. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesA conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Edge Computing Is Able to Prosper AI With Richer Data and Application Scenarios”
- Also stated in: § 2) Edge Computing Is Able to Prosper AI With Richer Data and Application Scenarios: · hedged
“edge computing is proposed to achieve low-latency data processing by sinking the computing capability from the cloud data center to the edge side, i.e., data generation source, which may enable AI processing with high performance.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Edge Computing Is Able to Prosper AI With Richer Data and Application Scenarios:
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Also stated in: § 2) Edge Computing Is Able to Prosper AI With Richer Data and Application Scenarios: · hedged
In the paper
“A flaw of [46] is that it does not consider unbalanced and nonindependent identical distribution (non-IID) data.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > C. Enabling Technologies
What the sources say
- Not checked — no full text of the sourcecites [46] 46. R. Shokri and V. Shmatikov, "Privacy-preserving deep learning," in Proc. 22nd ACM SIGSAC Conf. Comput. Commun. Secur., 2015, pp. 1310-1321. source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIn the paper
“big data have recently gone through a radical shift of data source from the megascale cloud datacenters to the increasingly widespread end devices, e.g., mobile devices and Internetof-Things (IoT) devices.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > I. I N T R O D U C T I O N
What the sources say
- Not matchedcites [3] 3. Cisco Global Cloud Index: Forecast and Methodology, 2016-2021, White Paper. [Online]. Available: https://www.cisco.com/c/en/us/ solutions/collateral/service-provider/global- cloud-index-gci/white-paper-c11-738085.html
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Not matched
- Probably not a substantial claim. 28
- It may state two separate claims as one. 27
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“VGG-16 [19] is trained to a deeper structure of 16 layers consisting of 13 convolution layers and three fully connected layers, requiring 138 million weights and 15.5G MACs to classify the image with a size of 224 × 224.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Convolution Neural Network:
What the sources say
- Partly supportedcites [19] 19. K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," 2014, arXiv:1409.1556. [Online]. Available: https://arxiv.org/abs/1409.1556 sourceExperimental — accuracy not yet measured
Table 1 (p. 3) identifies configuration D as a 16-weight-layer network with a 224 × 224 RGB input, and Table 2 (p. 3) lists 138 million parameters for D. Section 2.1 (p. 2) describes the three fully connected layers. Our copy does not report the claimed 15.5G MACs per image, so that part is unsupported. (LLM, conf=0.88)
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Partly supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“In particular, an update aggregation protocol named FedCS is developed to allow the centralized server to aggregate as many client updates as possible and to accelerate performance improvement in ML models.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Aggregation Frequency Control:
What the sources say
- Could not verifycites [52] 52. T. Nishio and R. Yonetani, "Client selection for federated learning with heterogeneous resources in mobile edge," 2018, arXiv:1804.08333. [Online]. Available: https://arxiv.org/abs/ 1804.08333 sourceExperimental — accuracy not yet measured
The Introduction (p. 2) describes pruning, quantization, and Huffman coding, while §6.3 (pp. 9–10) discusses inference speedups from compression. Searches for FedCS, client-update aggregation, and federated client selection found no passage addressing the claimed protocol or its effect on model improvement. (LLM, conf=0.76)
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Could not verify
Every source the paper cites for this claim was checked and supports it.
In the paper
“The optimization in [45] emphasizes the properties of unbalanced and non-IID since the distributed data may come from various sources.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > C. Enabling Technologies
What the sources say
- Supportedcites [45] 45. H. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. Y. Arcas, "Communication-efficient learning of deep networks from decentralized data," 2016, arXiv:1602.05629. [Online]. Available: https://arxiv.org/abs/1602.05629 source
The Introduction (p. 2) explicitly says the work emphasizes the non-IID and unbalanced properties of federated optimization. It explains that clients’ data reflect different users’ device usage and that users contribute varying amounts of data, supporting the citing sentence’s explanation of why those properties arise. (LLM, conf=0.94)
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Supported
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“when moving a tremendous amount of data across the wide area network (WAN), both monetary cost and transmission delay can be prohibitively high, and the privacy leakage can also be a major concern [5].”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > I. I N T R O D U C T I O N
What the sources say
- Could not verifycites [5] 5. Q. Pu et al., "Low latency geo-distributed data analytics," in Proc. ACM SIGCOMM, 2015, pp. 421-434. sourceExperimental — accuracy not yet measured
The Introduction and Section 6 (Advantages and disadvantages; page numbers unavailable) discuss generative models and their computational properties, not transferring large datasets over a WAN. Searches for WAN costs, transmission delay, and privacy leakage found no passage addressing the claim. (LLM, conf=0.75)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Could not verify
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“the physical proximity between the computing and information-generation sources promises several benefits compared to the traditional cloud-based computing paradigm, including low latency, energy efficiency, privacy protection, reduced bandwidth consumption, on-premises, and context awareness [6], [8].”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > I. I N T R O D U C T I O N
What the sources say
- Not checked — no full text of the sourcecites [6] 6. W. Shi, J. Cao, Q. Zhang, Y. Li, and L. Xu, "Edge computing: Vision and challenges," IEEE Internet Things J., vol. 3, no. 5, pp. 637-646, Oct. 2016. source
- Supportedcites [8] 8. Y. Mao, C. You, J. Zhang, K. Huang, and K. B. Letaief, "A survey on mobile edge computing: The communication perspective," IEEE Commun. Surveys Tuts., vol. 19, no. 4, pp. 2322-2358, 4th Quart., 2017. source
The comparison in “Mobile Edge Computing Versus Mobile Cloud Computing” (p. 3) identifies lower latency, mobile-device energy savings, context awareness, and improved privacy and security as MEC advantages. “Mobile Computing for 5G: From Clouds to Edges” (p. 1) explains that moving traffic toward the edge avoids burdening backhaul networks, while the privacy discussion (p. 3) gives enterprise-managed deployment as an example of keeping data on premises. (LLM, conf=0.88)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“A good performance improvement is obtained by transferring knowledge from both the intermediate layers and last layer of the teacher to a shallower student while other architectures and transfer techniques do not fare so well and some of them even lead to negative performance impact.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Not checked — no full text of the sourcecites [62] 62. R. Sharma, S. Biookaghazadeh, B. Li, and M. Zhao, "Are existing knowledge transfer techniques effective for deep learning with edge devices?" in Proc. IEEE Int. Conf. Edge Comput. (EDGE), Jul. 2018, pp. 42-49. source
- Could not verifycites [44] 44. Q. Chen, Z. Zheng, C. Hu, D. Wang, and F. Liu, "Data-driven task allocation for multi-task transfer learning on the edge," in Proc. IEEE 39th Int. Conf. Distrib. Comput. Syst. (ICDCS), 2019. sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 2 unverifiable. Expand the claim to see per-atom verdicts.
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
Every source the paper cites for this claim was checked and supports it.
In the paper
“The early work on random gossip algorithms is gossip averaging [63], which can fast converge toward a consensus among nodes by exchanging information peer to peer.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Supportedcites [63] 63. S. Boyd, A. Ghosh, B. Prabhakar, and D. Shah, "Randomized gossip algorithms," IEEE Trans. Inf. Theory, vol. 52, no. 6, pp. 2508-2530, Jun. 2006. source
The paper describes gossip averaging as nodes communicating with randomly chosen neighbors (Randomized Gossip Algorithms, p. 1). It studies convergence to the vector of averages and the design of the fastest gossip algorithm (Previous Results, p. 3; Randomized Gossip Algorithms, p. 1), supporting the peer-to-peer consensus paraphrase. (LLM, conf=0.89)
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Supported
In the paper
“The algorithm is based on the convergence analysis of distributed gradient descent and can be applied to federated learning in edge computing with provable convergence.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Aggregation Frequency Control:
What the sources say
- Not matchedcites [51] 51. S. Wang et al., "Adaptive federated learning in resource constrained edge computing systems," IEEE J. Sel. Areas Commun., vol. 37, no. 3, pp. 1205-1221, Jun. 2019.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As a result, the transmission latency of data offloading decreases, the data privacy increases, and the WAN bandwidth cost reduces.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 5) Level 4-Cloud-Edge Cotraining and Inference: Train-
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The main challenges of federated learning are optimization and communication.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > C. Enabling Technologies
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While edge computing and AI complement each other from a technical perspective, their application and popularization are also mutually beneficial.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Edge Computing Is Able to Prosper AI With Richer Data and Application Scenarios:
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Though the federated learning technique exploits a new decentralized deep learning architecture, it is built upon a central server for aggregating local updates.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > C. Enabling Technologies
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“As forecast by Gartner [33], more than 80% of enterprise IoT projects will include an AI component by 2022, up from only 10% today.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > A. Motivation and Benefits of Edge Intelligence
What the sources say
- Not matchedcites [33] 33. AI Trends for Enterprise Computing. [Online]. Available: https://www.gartner.com/ smarterwithgartner/3-ai-trends-for-enterprise- computing/
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This conflict indicates that there is no”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 5) Level 4-Cloud-Edge Cotraining and Inference: Train-
Load-bearingStrength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Energy efficiency is mainly affected by the size of the target training model and resources of the used devices.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 6) Energy Efficiency:
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
In the paper
“ResNet [20], the state-of-the-art effort, uses the”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Convolution Neural Network:
What the sources say
- Not checked — no full text of the sourcecites [20] 20. K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proc. IEEE CVPR, Jun. 2016, pp. 770-778. source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“real-time video analytics is envisioned to be a killer application for edge computing [12], [37], [38].”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) Edge Computing Can Be Popularized With AI Applications:
What the sources say
- Not checked — no full text of the sourcecites [12] 12. G. Ananthanarayanan et al., "Real-time video analytics: The killer app for edge computing," Computer, vol. 50, no. 10, pp. 58-67, 2017. source
- Not checked — no full text of the sourcecites [37] 37. H. Zhang, G. Ananthanarayanan, P. Bodik, M. Philipose, P. Bahl, and M. J. Freedman, "Live video analytics at scale with approximation and delay-tolerance," in Proc. USENIX NSDI, 2017, pp. 377-392.
- Not checked — no full text of the sourcecites [38] 38. C.-C. Hung et al., "VideoEdge: Processing camera streams using hierarchical clusters," in Proc. IEEE/ACM Symp. Edge Comput. (SEC), Oct. 2018, pp. 115-131. source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“AlexNet requires 61 million weights and 724 million MACs (multiply-add computation) to classify the image with a size of 227 × 227.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Convolution Neural Network:
What the sources say
- Not matchedcites [18] 18. A. Krizhevsky, I. Sutskever, and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," in Proc. NIPS, 2012, pp. 1097-1105.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
In the paper
“However, Gaia focuses on geodistributed datacenters that are capacity-unlimited, making it is not generally applicable to edge computing nodes whose capacity is highly constrained.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Aggregation Frequency Control:
What the sources say
- Not matchedcites [50] 50. K. Hsieh, A. Harlap, N. Vijaykumar, D. Konomis, G. R. Ganger, and P. B. Gibbons, "Gaia: Geo-distributed machine learning approaching LAN speeds," in Proc. NSDI, 2017, pp. 629-647.
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Not matched
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Under the decentralized training mode, the convergence value depends on the way the gradient is synchronized and updated.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 2) Convergence:
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“When training the DNN model by using the data originated at a massive of end devices, the raw data or intermediate data should be transferred out of the end devices.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 3) Privacy:
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Deep learning represents the state-of-the-art AI technology as well as a highly resource-demanding workload that naturally suits for edge computing.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > B. Deep Learning and Deep Neural Networks
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“When training the DNN model in a decentralized manner, both the computation and communication processes consume enormous energy.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 6) Energy Efficiency:
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.
- Rests on an assumption: For most end devices, they are energy-constrained.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, this is achieved at the cost of increased computational latency and energy consumption.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 5) Level 4-Cloud-Edge Cotraining and Inference: Train-
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“GoogleNet achieves a better accuracy performance than VGG-16, while only requiring seven million weights and 1.43G MACs to process the image with the same size.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Convolution Neural Network:
What the sources say
- Not checked — no full text of the sourcecites [28] 28. C. Szegedy et al., "Going deeper with convolutions," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., Jun. 2015, pp. 1-9. source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
In the paper
“MeDNN improves the DNN model inference by 1.86-2.44×with two to four worker nodes and saves 26.5% of additional computing time and 14.2% of extra communication time.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Model Compression:
What the sources say
- Not checked — no full text of the sourcecites [84] 84. J. Mao et al., "MeDNN: A distributed mobile system with enhanced partition and deployment for large-scale DNNs," in Proc. 36th Int. Conf. Comput.-Aided Design, Nov. 2017, pp. 751-756. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Communication overhead is affected by the size of the original input data, the way of transmission, and the available bandwidth.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > VII. C O N C L U S I O N
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Due to the combinatorial nature, the high-dimensional configuration problem involves a huge search space of parameters and is very challenging to tackle.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > D. Summary of Existing Systems and Frameworks
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Rests on an assumption: Standard performance indicators such as top-k accuracy or mean average precision fail to reflect the runtime performance of DNN model inference on edge devices.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, data parallelism may bring heavy overhead of communication, while model parallelism usually leads to severe underutilization of computation resources.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearIt may state two separate claims as one.
In the paper
“PipeDream shows the advantage of reducing communication overhead and utilizing computing resource efficiently.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Not checked — no full text of the sourcecites [61] 61. A. Harlap, D. Narayanan, A. Phanishayee, V. Seshadri, G. R. Ganger, and P. B. Gibbons, "PipeDream: Fast and efficient pipeline parallel DNN training," 2018, arXiv:1806.03377. [Online]. Available: https://arxiv.org/abs/1806.03377 source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The communication latency may vary from the size of transmitted raw or intermediate data and the bandwidth of network connection.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > VII. C O N C L U S I O N
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“By deployment on large-scale systems, Daily et al. [66] show that the trivial gossip-based algorithms at scale lead to a communication imbalance, poor convergence, and heavy communication overhead.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Not checked — no full text of the sourcecites [66] 66. J. Daily, A. Vishnu, C. Siegel, T. Warfel, and V. Amatya, "GossipGraD: Scalable deep learning using gossip communication based asynchronous gradient descent," 2018, arXiv:1803.05880. [Online]. Available: https://arxiv.org/ abs/1803.05880 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“FoggyCache reduces computation latency and energy consumption by a factor of 3× to 10×.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) Edge Caching:
What the sources say
- Not checked — no full text of the sourcecites [95] 95. P. Guo, B. Hu, R. Li, and W. Hu, "FoggyCache: Cross-device approximate computation reuse," in Proc. ACM Mobicom, 2018, pp. 19-34. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Arguably, latency is one of the most fundamental performance indicators of distributed DNN model training since it directly influences when the trained model is available for use.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > VII. C O N C L U S I O N
Strength:hedged· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“This is based on the important observation that a DNN model can be split inside between two successive layers with two partitions deployed on different locations without losing accuracy.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
In the paper
“the proposed automatic optimization framework AdaDeep [75] systematically formulates the goals and constraints on accuracy, latency, storage, and energy into a unified optimization problem and leverages DRL to effectively find a good combination of compression techniques.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Model Compression:
What the sources say
- Not checked — no full text of the sourcecites [75] 75. S. Liu, Y. Lin, Z. Zhou, K. Nan, H. Liu, and J. Du, "On-demand deep model compression for mobile devices: A usage-driven model selection framework," in Proc. 16th Annu. Int. Conf. Mobile Syst., Appl., Services, 2018, pp. 389-400. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The decentralized architecture is naturally friendly to user's privacy, for which the systems that are based on the decentralized architecture such as BlockFL and GossipGraD typically preserve privacy better.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > Table 2 Overview of Systems and Frameworks on EI Model Training
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Intuitively, this communication overhead increases the training latency, energy, and bandwidth consumption.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > VII. C O N C L U S I O N
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“Precog uses predictions of Markov chains to prefetch data onto the mobile device and reach a speedup of 5×.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) Edge Caching:
What the sources say
- Not checked — no full text of the sourcecites [94] 94. U. Drolia, K. Guo, and P . Narasimhan, "Precog: Prefetching for image recognition applications at the edge," in Proc. ACM/IEEE Symp. Edge Comput. (SEC), Oct. 2017, p. 17. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Compared with the DNN training under the cloud-based framework, the DNN training under edge-based framework pays more attention to protecting users' privacy and training an available deep learning model faster.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > Table 2 Overview of Systems and Frameworks on EI Model Training
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The latency of the distributed training process typically consists of the computation latency and the communication latency.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > VII. C O N C L U S I O N
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“Both Blot et al. [64] and Jin et al. [65] apply gossip algorithms on the updates of SGD, but neither of them has performance convergence degradation at large scale.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
What the sources say
- Not checked — no full text of the sourcecites [64] 64. M. Blot, D. Picard, M. Cord, and N. Thome, "Gossip training for deep learning," 2016, arXiv:1611.09726. [Online]. Available: https://arxiv.org/abs/1611.09726 source
- Not checked — no full text of the sourcecites [65] 65. P. H. Jin, Q. Yuan, F. Iandola, and K. Keutzer, "How to scale distributed deep learning?" 2016, arXiv:1611.04581. [Online]. Available: https://arxiv.org/abs/1611.04581 source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
In the paper
“ReXCam reduces computation workload by 4.6× and improves DNN model inference accuracy by 27%.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 5) Input Filtering:
What the sources say
- Not checked — no full text of the sourcecites [99] 99. S. Jain, J. Jiang, Y. Shu, G. Ananthanarayanan, and J. Gonzalez, "ReXCam: Resource-efficient, cross-camera video analytics at enterprise scale," 2018, arXiv:1811.01268. [Online]. Available: https://arxiv.org/abs/1811.01268 source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“For the object recognition application, Huang et al. [120] investigate the influence of the main factors, e.g., number of proposals, input image size, and the selection of feature extractor, on inference speed and accuracy.Based on their experiment results, a new combination of these factors is found to outperform the state-of-the-art method.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > D. Tradeoff Design With Various DNN Performance Metrics
What the sources say
- Not checked — no full text of the sourcecites [120] 120. J. Huang et al., "Speed/accuracy trade-offs for modern convolutional object detectors," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Jul. 2017, pp. 7310-7311. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Rests on an assumption: Standard performance indicators such as top-k accuracy or mean average precision fail to reflect the runtime performance of DNN model inference on edge devices.
In the paper
“Compared to the approach that uploads the entire model, IONN significantly improves query performance and energy consumption during DNN model uploading.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 1) Model Compression:
What the sources say
- Not checked — no full text of the sourcecites [82] 82. H.-J. Jeong, H.-J. Lee, C. H. Shin, and S.-M. Moon, "IONN: Incremental offloading of neural network computations from mobile devices to edge servers," in Proc. ACM Symp. Cloud Comput., 2018, pp. 401-411. source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The gossip distributed algorithms enjoy the advantage of full asynchronization and total decentralization as they have no requirement on centralized nodes or variables.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“the cascading network [91] simply adds the MP layer and fully connected layer to the standard DNN model and achieves a speedup of 20%.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > The aggregation methods work when multiple mobile
What the sources say
- Not checked — no full text of the sourcecites [91] 91. S. Leroux et al., "The cascading neural network: Building the Internet of smart things," Knowl. Inf. Syst., vol. 52, no. 3, pp. 791-814, 2017.
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“This process will tend to work if the features are general (i.e., suitable to both base and target tasks) instead of specific to the base task.”
§ Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing This article conducts a comprehensive survey of the research efforts on edge intelligence. It provides an overview of the architectures, frameworks, and emerging key > 4) DNN Splitting:
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper relies on this without support.
There is a lack of a dedicated venue for summarizing, discussing, and disseminating recent advances of edge intelligence in both industry and academia.
“There is, in general, a lack of venue dedicated for summarizing, discussing, and disseminating the recent advances of EI, in both industrial and academia.”
The paper relies on this without support.
Research and practice on edge intelligence are still in a very early stage.
“Notably, research and practice on this emerging interdiscipline-EI-are still in a very early stage.”
The paper relies on this without support.
The edge ecosystem will present many novel application scenarios for AI and fuel the continuous booming of AI.
“Clearly, via bringing the huge volumes of data to AI, the edge ecosystem will present many novel application scenarios for AI and fuel the continuous booming of AI.”
The paper relies on this without support.
For most end devices, they are energy-constrained.
“However, for most end devices, they are energy-constrained.”
The paper relies on this without support.
It is inevitable to deal with privacy issues when raw or intermediate data are transferred out of end devices.
“Obviously, it is inevitable to deal with privacy issues in this scenario.”
The paper relies on this without support.
For some real-time intelligent mobile applications, they usually have stringent deadline requirement such as 100-ms latency.
“For some real-time intelligent mobile applications (e.g., AR/VR mobile gaming and intelligent robots), they usually have stringent deadline requirement such as 100-ms latency.”
The paper relies on this without support.
For some mobile applications requiring high reliability, they demand ultrahigh accuracy on DNN model inference.
“For some mobile applications requiring a high level of reliability, such as self-driving car and face authentication, they demand the ultrahigh accuracy on the DNN model inference.”
The paper relies on this without support.
Most existing efforts use a single compression technique.
“While most existing efforts use a single compression technique, they may not suffice to meet the diverse requirements and constraints on accuracy, latency, storage, and energy imposed by some IoT devices.”
The paper relies on this without support.
Standard performance indicators such as top-k accuracy or mean average precision fail to reflect the runtime performance of DNN model inference on edge devices.
“However, it is difficult for software developers to choose an appropriate DNN model for the EI application because the standard performance indicators such as top-k accuracy or mean average precision fail to reflect the runtime performance of DNN model inference on edge devices.”
The paper relies on this without support.
Most current edge AI programming/software platforms mainly serve as relays for connecting to cloud data centers.
“Nevertheless, currently, most of these platforms mainly serve as relays for connecting to the powerful cloud data centers.”
The paper relies on this without support.
Most deep learning-based AI models are highly resource-intensive.
“Most of the deep learning-based AI models are highly resourceintensive, which means that powerful computing capability supported by abundant hardware resources (e.g., GPU, field-programmable gate array (FPGA), and TPU) is important to boost the performance of these AI models.”
The paper relies on this without support.
For an EI application with a specific mission, there is usually a series of DNN model candidates that are capable of finishing the task.
“For an EI application with a specific mission, there is usually a series of DNN model candidates that are capable of finishing the task.”
The paper relies on this without support.
The service environments for many EI application scenarios are of high dynamics and it is hard to accurately predict future events.
“Since for many EI application scenarios (e.g., smart cities), the service environments are of high dynamics and it is hard to accurately predict future events.”
The paper relies on this without support.
The high-efficiency operation of EI services may require close collaboration and integration across different service providers.
“The high-efficiency operation of EI services may require close collaboration and integration across different service providers, e.g., for implementing expanded resource sharing and smooth service handover.”
The paper relies on this without support.
A user can be a service consumer and, meanwhile, a data generator as well.
“Also, for EI service, a user can be a service consumer and, meanwhile, a data generator as well.”
The paper relies on this without support.
There is a stringent need to push the AI frontier from the cloud to the network edge.
“Driving by flourishing of both AI and IoT, there is a stringent need to pushing the AI frontier from the cloud to the network edge.”
The paper relies on this without support.
Edge computing is widely recognized as a promising solution to support computation-intensive AI applications in resource-constrained environments.
“To fulfill this trend, edge computing has been widely recognized as a promising solution to support computation-intensive AI applications in resource-constrained environments.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Additionally, UAVs generate lots of data at the network edge and directly sharing original data sets between many edge nodes has a high risk associated with it [8].”
From A Survey on the Convergence of Edge Computing and AI for UAVs: Opportunities and Challenges (2022) · cites this paper as [8]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Edge Intelligence (EI) Zhou et al. (2019) proposes a new paradigm combining AI and edge computing, with the goal of performing distributed computing of DL models.”
From Deepfakes: current and future trends (2024) · cites this paper as Zhou et al. 2019
The cited source says what the claim attributes to it.
In the citing paper
“This entails some advantages such as low latency, energy and bandwidth efficiency, privacy protection, and context awareness (Zhou et al. 2019).”
From Deepfakes: current and future trends (2024) · cites this paper as Zhou et al. 2019
The cited source says what the claim attributes to it.
In the citing paper
“In the case of training a deep learning model in an edge computing environment, where distributed models are trained locally first and then updates are aggregated centrally (e.g., federated learning), the control of updates aggregation frequency has a significant influence on the communication overhead.”
From A Survey on the Convergence of Edge Computing and AI for UAVs: Opportunities and Challenges (2022) · cites this paper as [8]
In the citing paper
“Edge AI provides a promising solution for connected intelligence by enabling data collection, processing, transmission, and consumption at the network edge [29], [30].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [30]
In the citing paper
“All of these reviews only have a small fraction to discuss the edge DNN design which is the key element of edge intelligence systems.”
From Bringing ai to edge: from deep learning’s perspective (2022) · cites this paper as [13]
In the citing paper
“Smart city [113] is another important application scenario for deep-learning-enabled edge computing.”
From Deep learning for edge computing applications: A State-of-the-Art survey (2020) · cites this paper as [113]
In the citing paper
“In the current scenario, Blockchain provides a high degree of privacy for health data [49].”
From 6G Communication Technology: A Vision on Intelligent Healthcare (2021) · cites this paper as [49]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Inference metrics are counted as five but six are listed
MinorThe section introduces five inference metrics, but its numbered list runs from latency through memory footprint and contains six.
“To describe the service quality of the EI model inference, we introduce the following five metrics.”
“## 6) Memory Footprint:”
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148 · cited 5×
“GoogleNet achieves a better accuracy performance than VGG-16, while only requiring seven million weights and 1.43G MACs to process the image with the same size.”
BackgroundMethodInfluentialImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 7×
“This is because for DNNs as exemplified by AlexNet, the energy of the convolutional layers dominates the total energy cost, while the number in the fully connected layers contributes most of the total number of weights in the DNN.”
BackgroundMethodInfluentialHochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 4×
“4, each neuron in LSTM is called memory cell and includes a multiplicative forget gate, input gate, and output gate.”
BackgroundInfluentialLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562 · cited 2×
“GoogleNet achieves a better accuracy performance than VGG-16, while only requiring seven million weights and 1.43G MACs to process the image with the same size.”
MethodLearning representations by back-propagating errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986) · Nature · doi:10.1038/323533a0 · Cited by 31,892 · cited 1×
“Leveraging the backpropagation mechanism [26], [27], the error rate is propagated back across the whole neural network, and the weights are updated based on the gradient and the learning rate.”
(2016) · Cited by 31,590
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357 · cited 2×
“[32] V.”
SSD: Single Shot MultiBox Detector
Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.-Y., & Berg, A. C. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46448-0_2 · Cited by 22,128 · cited 3×
“[24] W. Liu et al., “SSD: Single shot multibox detector,” in Proc.”
Labaca-Castro, R. (2023) · Springer eBooks · doi:10.1007/978-3-658-40442-0_9 · Cited by 21,119
GAN（Generative Adversarial Nets）
淳司, 柴. (2017) · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics · doi:10.3156/jsoft.29.5_177_2 · Cited by 20,507
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., & Adam, H. (2017) · arXiv · doi:10.48550/arxiv.1704.04861 · Cited by 19,015 · cited 4×
“[21] A. G. Howard et al., “MobileNets: Efficient convolutional neural networks for mobile vision applications,” 2017, arXiv:1704.04861.”
BackgroundInfluentialCommunication-Efficient Learning of Deep Networks from Decentralized Data
McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Arcas, B. A. Y. (2016) · arXiv · doi:10.48550/arxiv.1602.05629 · Cited by 14,215 · cited 4×
“[45] propose to increase the computation of local updates on clients.”
BackgroundMethodInfluentialEdge computing: Vision and challenges
Shi, W., Cao, J., Zhang, Q., Li, Y., & Xu, L. (2016) · IEEE Internet of Things Journal · doi:10.1109/jiot.2016.2579198 · Cited by 8,733 · cited 2×
“Essentially, the physical proximity between the computing and information-generation sources promises several benefits compared to the traditional cloud-based computing paradigm, including low latency, energy efficiency, privacy protection, reduced bandwidth consumption, on-premises, and context awareness [6], [8].”
BackgroundNatural Language Processing (almost) from ScratchPreprint
Collobert, R., Weston, J., Bottou, L., Karlen, M., Kavukcuoglu, K., & Kuksa, P. P. (2011) · arXiv (Cornell University) · doi:10.48550/arxiv.1103.0398 · Cited by 6,923 · cited 1×
“MLP models are the most basic DNN, which is composed of a series of fully connected layers [17].”
BackgroundLarge-Scale Machine Learning with Stochastic Gradient Descent
Bottou, L. (2010) · Proceedings of COMPSTAT'2010 · doi:10.1007/978-3-7908-2604-3_16 · Cited by 5,795 · cited 25×
“Amiri and Gunduz [57] implement DSGD at the wireless edge with the help of a remote parameter server and further develop DSGD in digital and analog schemes, respectively.”
MethodInfluentialNeural Architecture Search with Reinforcement LearningPreprint
Zoph, B., & Le, Q. V. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.01578 · Cited by 5,705 · cited 1×
“Instead of utilizing the existing resource-intensive AI models, we can leverage the AutoML idea [116] and the neural architecture search (NAS) techniques [117] to devise resource-efficient edge AI models tailored to the hardware resource constraints of the underlying edge devices and servers.”
MethodA Survey on Mobile Edge Computing: The Communication Perspective
Mao, Y., You, C., Zhang, J., Huang, K., & Letaief, K. B. (2017) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2017.2745201 · Cited by 5,607 · cited 1×
“Essentially, the physical proximity between the computing and information-generation sources promises several benefits compared to the traditional cloud-based computing paradigm, including low latency, energy efficiency, privacy protection, reduced bandwidth consumption, on-premises, and context awareness [6], [8].”
BackgroundSelf-citationBackpropagation through time: what it does and how to do it
Werbos, P. J. (1990) · Proceedings of the IEEE · doi:10.1109/5.58337 · Cited by 4,879 · cited 2×
“The training of RNN Vol. 107, No. 8, August 2019 | PROCEEDINGS OF THE IEEE 1741 is based on backpropagation through time (BPTT) [29].”
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1510.00149 · Cited by 4,662 · cited 2×
“The follow-up work deep compression [71], which blends the advantages of pruning, weight sharing, and Huffman coding to compress DNNs, further pushes the compression ratio to 35–49×.”
BackgroundMethodEfficient Processing of Deep Neural Networks: A Tutorial and Survey
Sze, V., Chen, Y.-H., Yang, T.-J., & Emer, J. S. (2017) · Proceedings of the IEEE · doi:10.1109/jproc.2017.2761740 · Cited by 4,215 · cited 1×
“Interested readers are encouraged to refer to the recent monograph [114] for more comprehensive discussions about recent advancements on hardware acceleration for DNN processing.”
BackgroundDeep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196
Federated Learning: Strategies for Improving Communication Efficiency
Konečný, J., McMahan, H. B., Yu, F. X., Richtárik, P., Suresh, A. T., & Bacon, D. (2016) · arXiv · doi:10.48550/arxiv.1610.05492 · Cited by 3,766 · cited 1×
“[47] propose to reduce communication cost with two new update schemes, namely, structured update and sketched update.”
BackgroundThe case for vm-based cloudlets in mobile computing
Satyanarayanan, M., Bahl, P., Caceres, R., & Davies, N. (2009) · IEEE Pervasive Computing · doi:10.1109/mprv.2009.82 · Cited by 3,660 · cited 1×
“To clear up the doubt, Microsoft has conducted continuous exploration on what kinds should be moved from the cloud to the edge since 2009 [35], ranging from voice command recognition, AR/VR, and interactive cloud gaming [36] to real-time video analytics.”
BackgroundDeep Learning: Methods and Applications
Deng, L., & Yu, D. (2014) · Foundations and Trends® in Signal Processing · doi:10.1561/2000000039 · Cited by 3,480 · cited 2×
“, AlphaGo) and robotics [2].”
Speed/Accuracy Trade-Offs for Modern Convolutional Object Detectors
Huang, J., Rathod, V., Sun, C., Zhu, M., Korattikara, A., Fathi, A., Fischer, I., Wojna, Z., Song, Y., Guadarrama, S., et al. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.351 · Cited by 2,743 · cited 1×
“[120] investigate the influence of the main factors, e.”
BackgroundBoyd, S., Ghosh, A., Prabhakar, B., & Shah, D. (2006) · IEEE Transactions on Information Theory · doi:10.1109/tit.2006.874516 · Cited by 2,439 · cited 1×
“The early work on random gossip algorithms is gossip averaging [63], which can fast converge toward a consensus among nodes by exchanging information peer to peer.”
BackgroundPrivacy-preserving deep learning
Shokri, R., & Shmatikov, V. (2015) · Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/2810103.2813687 · Cited by 2,356 · cited 3×
“Shokri and Shmatikov [46] design a selective SGD (SSGD) protocol, allowing the clients to train independently on their own data sets and selectively share small subsets of their model’s key parameters to the centralized aggregator.”
BackgroundInfluentialAdaptive Federated Learning in Resource Constrained Edge Computing Systems
Wang, S., Tuor, T., Salonidis, T., Leung, K. K., Makaya, C., He, T., & Chan, K. (2019) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2019.2904348 · Cited by 2,232 · cited 1×
“[51] propose a control algorithm that determines the best tradeoff between the local update and global parameter aggregation under a given resource budget.”
BackgroundClient Selection for Federated Learning with Heterogeneous Resources in Mobile Edge
Nishio, T., & Yonetani, R. (2018) · arXiv · doi:10.48550/arxiv.1804.08333 · Cited by 1,875
Client Selection for Federated Learning with Heterogeneous Resources in Mobile Edge
Nishio, T., & Yonetani, R. (2019) · ICC 2019 - 2019 IEEE International Conference on Communications (ICC) · doi:10.1109/icc.2019.8761315 · Cited by 1,875
Deep Learning: Methods and Applications
Deng, L. (2014) · now publishers, Inc. eBooks · doi:10.1561/9781601988157 · Cited by 1,789
Introduction to multi-layer feed-forward neural networks
Svozil, D., Kvasnicka, V., & Pospichal, J. (1997) · Chemometrics and Intelligent Laboratory Systems · doi:10.1016/s0169-7439(97)00061-0 · Cited by 1,600 · cited 1×
“Among the existing ML methods, deep learning, by leveraging artificial neural networks (ANNs) [16] to learn the deep representation of the data, has resulted in an amazing performance in multiple tasks, including image classification, face recognition, and so on.”
MethodGoing Deeper with Convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2014) · arXiv · doi:10.48550/arxiv.1409.4842 · Cited by 1,461
AMC: AutoML for Model Compression and Acceleration on Mobile Devices
He, Y., Lin, J., Liu, Z., Wang, H., Li, L.-J., & Han, S. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01234-2_48 · Cited by 1,400 · cited 1×
“Instead of utilizing the existing resource-intensive AI models, we can leverage the AutoML idea [116] and the neural architecture search (NAS) techniques [117] to devise resource-efficient edge AI models tailored to the hardware resource constraints of the underlying edge devices and servers.”
MethodBranchyNet: Fast inference via early exiting from deep neural networks
Teerapittayanon, S., McDanel, B., & Kung, H. T. (2016) · 2016 23rd International Conference on Pattern Recognition (ICPR) · doi:10.1109/icpr.2016.7900006 · Cited by 1,326 · cited 1×
“BranchyNet [87] is a programming framework that implements the model early-exit mechanism.”
MethodEyeriss: a spatial architecture for energy-efficient dataflow for convolutional neural networks
Chen, Y.-H., Emer, J., & Sze, V. (2016) · ACM SIGARCH Computer Architecture News · doi:10.1145/3007787.3001177 · Cited by 1,182 · cited 1×
“However, for energy-constrained end devices, the above magnitude-based weight pruning method may not be directly applicable since empirical measurements show that the reduction in the number of weights does not necessarily translate into significant energy saving [72].”
BackgroundDeep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training
Lin, Y., Han, S., Mao, H., Wang, Y., & Dally, W. J. (2017) · arXiv · doi:10.48550/arxiv.1712.01887 · Cited by 1,134 · cited 1×
“Inspired by [53], Tao and Li [54] propose edge SGD (eSGD), a family of sparse schemes with both convergence and practical performance guarantees.”
MethodIn-Edge AI: Intelligentizing Mobile Edge Computing, Caching and Communication by Federated Learning
Wang, X., Han, Y., Wang, C., Zhao, Q., Chen, X., & Chen, M. (2019) · IEEE Network · doi:10.1109/mnet.2019.1800286 · Cited by 1,030 · cited 1×
“Indeed, the marriage of edge computing and AI has given rise to a new research area, namely, “edge intelligence (EI)” or “edge AI” [9], [10].”
BackgroundSelf-citationBlockchained On-Device Federated Learning
Kim, H., Park, J., Bennis, M., & Kim, S.-L. (2018) · arXiv · doi:10.48550/arxiv.1808.03949 · Cited by 882
Distributed Deep Neural Networks Over the Cloud, the Edge and End Devices
Teerapittayanon, S., McDanel, B., & Kung, H. T. (2017) · 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs.2017.226 · Cited by 843 · cited 4×
“DDNNs [88] N/A Cloud-Edge-Device Level-1 • Latency • Accuracy • Model selection N/A • Latency reduction: over 20×”
MethodInfluentialDesigning Energy-Efficient Convolutional Neural Networks Using Energy-Aware Pruning
Yang, T.-J., Chen, Y.-H., & Sze, V. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.643 · Cited by 732 · cited 1×
“Based on this energy estimation tool, an energy-aware pruning method called EAP [73] is proposed.”
MethodNeurosurgeon: Collaborative Intelligence Between the Cloud and Mobile Edge
Kang, Y., Hauswald, J., Gao, C., Rovinski, A., Mudge, T., Mars, J., & Tang, L. (2017) · Proceedings of the Twenty-Second International Conference on Architectural Support for Programming Languages and Operating Systems · doi:10.1145/3037697.3037698 · Cited by 673 · cited 3×
“For the model partition between the server and the device, Neurosurgeon [79] represents an iconic effort.”
BackgroundInfluentialEyeriss: A Spatial Architecture for Energy-Efficient Dataflow for Convolutional Neural Networks
Chen, Y., Emer, J., & Sze, V. (2016) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca.2016.40 · Cited by 629
Machine Learning at the Wireless Edge: Distributed Stochastic Gradient Descent Over-the-Air
Amiri, M., & Gündüz, D. (2019) · International Symposium on Information Theory · doi:10.1109/isit.2019.8849334 · Cited by 608
Machine Learning at the Wireless Edge: Distributed Stochastic Gradient Descent Over-the-Air
Mohammadi Amiri, M., & Gunduz, D. (2020) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2020.2981904 · Cited by 548
Large-Scale Machine Learning with Stochastic Gradient Descent
Bottou, L. (2011) · Chapman & Hall/CRC series in computer science & data analysis/Series in computer science and data analysis · doi:10.1201/b11429-4 · Cited by 536
Stich, S. U., Cordonnier, J.-B., & Jaggi, M. (2018) · arXiv · doi:10.48550/arxiv.1809.07599 · Cited by 508 · cited 1×
“A concise convergence analysis of sparsified SGD is given in [55], where SGD is analyzed with k-sparsification or compression (e.”
MethodDeepThings: Distributed Adaptive Deep Learning Inference on Resource-Constrained IoT Edge Clusters
Zhao, Z., Barijough, K. M., & Gerstlauer, A. (2018) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2018.2858384 · Cited by 502 · cited 2×
“[86] Z. Zhao, K. M. Barijough, and A. Gerstlauer, “DeepThings: Distributed adaptive deep learning inference on resource-constrained IoT edge clusters,” IEEE Trans.”
MethodReal-Time Video Analytics: The Killer App for Edge Computing
Ananthanarayanan, G., Bahl, P., Bodik, P., Chintalapudi, K., Philipose, M., Ravindranath, L., & Sinha, S. (2017) · Computer · doi:10.1109/mc.2017.3641638 · Cited by 496 · cited 2×
“By comparison, real-time video analytics is envisioned to be a killer application for edge computing [12], [37], [38].”
BackgroundTowards wearable cognitive assistance
Ha, K., Chen, Z., Hu, W., Richter, W., Pillai, P., & Satyanarayanan, M. (2014) · Proceedings of the 12th annual international conference on Mobile systems, applications, and services · doi:10.1145/2594368.2594383 · Cited by 476 · cited 1×
“These efforts have boosted a wide spectrum of AI applications, spanning from live video analytics [12], cognitive assistance [13] to precision agriculture, smart home [14], and Industrial Internet of Things (IIoT) [15].”
BackgroundMachine Learning at Facebook: Understanding Inference at the Edge
Wu, C.-J., Brooks, D. M., Chen, K., Chen, D., Choudhury, S., Dukhan, M., Hazelwood, K., Isaac, E., Jia, Y., Jia, B., et al. (2019) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca.2019.00048 · Cited by 473 · cited 1×
“On the other hand, unlike high-performance discrete GPUs on the cloud data center, there is no dedicated high-bandwidth memory for mobile GPUs on mobile devices [69].”
BackgroundChameleon: scalable adaptation of video analytics
Jiang, J., Ananthanarayanan, G., Bodik, P., Sen, S., & Stoica, I. (2018) · Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication · doi:10.1145/3230543.3230574 · Cited by 467
Deep Learning for Smart Industry: Efficient Manufacture Inspection System With Fog Computing
Li, L., Ota, K., & Dong, M. (2018) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2018.2842821 · Cited by 466 · cited 2×
“DeepIns [15] proposes a manufacture inspection system for the smart industry using DNN model early exit.”
BackgroundMethodDeepDecision: A Mobile Deep Learning Framework for Edge Video Analytics
Ran, X., Chen, H., Zhu, X., Liu, Z., & Chen, J. (2018) · IEEE INFOCOM 2018 - IEEE Conference on Computer Communications · doi:10.1109/infocom.2018.8485905 · Cited by 450 · cited 1×
“DeepDecision [112] formulates the knob-tunning problem as a multiple-choice multiple-constraint knapsack program and solves it with an improved brute-force search method.”
MethodDeepX: A Software Accelerator for Low-Power Deep Learning Inference on Mobile Devices
Lane, N. D., Bhattacharya, S., Georgiev, P., Forlivesi, C., Jiao, L., Qendro, L., & Kawsar, F. (2016) · 2016 15th ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN) · doi:10.1109/ipsn.2016.7460664 · Cited by 450 · cited 3×
“Model Compression • Weight pruning and quantization to reduce storage and computation [70]–[77]”
BackgroundMethodInfluentialMinerva: enabling low-power, highly-accurate deep neural network accelerators
Reagen, B., Whatmough, P., Adolf, R., Rama, S., Lee, H., Lee, S. K., Hernández-Lobato, J. M., Wei, G.-Y., & Brooks, D. (2016) · ACM SIGARCH Computer Architecture News · doi:10.1145/3007787.3001165 · Cited by 446 · cited 2×
“[74] B. Reagen et al., “Minerva: Enabling low-power, highly-accurate deep neural network accelerators,” ACM SIGARCH Comput.”
Gradient Coding: Avoiding Stragglers in Distributed Learning
Tandon, R., Lei, Q., Dimakis, A. G., & Karampatziakis, N. (2017) · International Conference on Machine Learning · Cited by 431
Dynamic Adaptive DNN Surgery for Inference Acceleration on the Edge
Hu, C., Bao, W., Wang, D., & Liu, F. (2019) · IEEE INFOCOM 2019 - IEEE Conference on Computer Communications · doi:10.1109/infocom.2019.8737614 · Cited by 416 · cited 1×
“[80] propose an approximation algorithm that provides worst-case performance guarantee, based on the graph min-cut method.”
Method(2013) · doi:10.4324/9780203763247 · Cited by 400 · cited 1×
“Leveraging the backpropagation mechanism [26], [27], the error rate is propagated back across the whole neural network, and the weights are updated based on the gradient and the learning rate.”
Edge Intelligence: On-Demand Deep Learning Model Co-Inference with Device-Edge Synergy
Li, E., Zhou, Z., & Chen, X. (2018) · Proceedings of the 2018 Workshop on Mobile Edge Communications · doi:10.1145/3229556.3229562 · Cited by 385 · cited 2×
“Indeed, the marriage of edge computing and AI has given rise to a new research area, namely, “edge intelligence (EI)” or “edge AI” [9], [10].”
BackgroundMethodSelf-citationNoScope: optimizing neural network queries over video at scale
Kang, D., Emmons, J., Abuzaid, F., Bailis, P., & Zaharia, M. (2017) · Proceedings of the VLDB Endowment · doi:10.14778/3137628.3137664 · Cited by 367
Glimpse: Continuous, Real-Time Object Recognition on Mobile Devices
Chen, T. Y.-H., Ravindranath, L., Deng, S., Bahl, P., & Balakrishnan, H. (2015) · Proceedings of the 13th ACM Conference on Embedded Networked Sensor Systems · doi:10.1145/2809695.2809711 · Cited by 358 · cited 2×
“inference with the model of full precision. &DFKHG 5HVXOWV 0RGHO ZLWKIXOO SUHFLVLRQ 0RELOH'HYLFHV (GJH6HUYHU &ORXG'DWDFHQWHU Fig. 15: The process of semantic cache technique Glimpse [92] is a pioneering effort to introduce cache technique to DNN inference task. For an object detection application, Gl…”
BackgroundMethodMoDNN: Local distributed mobile computing system for Deep Neural Network
Mao, J., Chen, X., Nixon, K. W., Krieger, C., & Chen, Y. (2017) · Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017 · doi:10.23919/date.2017.7927211 · Cited by 331 · cited 6×
“Note that DNN layers are partitioned horizontally in MoDNN and MeDNN, whereas DeepThings [86] employs a fused tile partitioning method that partitions the DNN layers vertically to reduce the memory footprint.”
MethodInfluentialSelf-citation(2017) · Networked Systems Design and Implementation · Cited by 329
Natural Language Processing (Almost) from Scratch
CollobertRonan, WestonJason, BottouLéon, KarlenMichael, KavukcuogluKoray, & KuksaPavel (2011) · Journal of Machine Learning Research · doi:10.5555/1953048.2078186 · Cited by 309
NestDNN: Resource-Aware Multi-Tenant On-Device Deep Learning for Continuous Mobile Vision
Fang, B., Zeng, X., & Zhang, M. (2018) · Proceedings of the 24th Annual International Conference on Mobile Computing and Networking · doi:10.1145/3241539.3241559 · Cited by 291 · cited 6×
“Taking the dynamics of runtime resources into consideration, NestDNN [107] is proposed to offering flexible resource–accuracy tradeoffs for each DNN model.”
MethodInfluentialVideoEdge: Processing Camera Streams using Hierarchical Clusters
Hung, C.-C., Ananthanarayanan, G., Bodik, P., Golubchik, L., Yu, M., Bahl, P., & Philipose, M. (2018) · 2018 IEEE/ACM Symposium on Edge Computing (SEC) · doi:10.1109/sec.2018.00016 · Cited by 287 · cited 1×
“By comparison, real-time video analytics is envisioned to be a killer application for edge computing [12], [37], [38].”
BackgroundMinerva: Enabling Low-Power, Highly-Accurate Deep Neural Network Accelerators
Reagen, B., Whatmough, P. N., Adolf, R., Rama, S., Lee, H., Lee, S. K., Hernández-Lobato, J. M., Wei, G.-Y., & Brooks, D. M. (2016) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca.2016.32 · Cited by 275
Notice of Removal: A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics
Osia, S. A., Shamsabadi, A. S., Sajadmanesh, S., Taheri, A., Katevas, K., Rabiee, H. R., Lane, N. D., & Haddadi, H. (2020) · IEEE Internet of Things Journal · doi:10.1109/jiot.2020.2967734 · Cited by 269
Adaptive Neural Networks for Efficient Inference
Bolukbasi, T., Wang, J., Dekel, O., & Saligrama, V. (2017) · arXiv · doi:10.48550/arxiv.1702.07811 · Cited by 251 · cited 1×
“[89] train a policy that determines whether the current samples should proceed to the next layer by adding regularization to the evaluation latency of the DNN model.”
BackgroundLow Latency Geo-distributed Data Analytics
Pu, Q., Ananthanarayanan, G., Bodik, P., Kandula, S., Akella, A., Bahl, P., & Stoica, I. (2015) · Proceedings of the 2015 ACM Conference on Special Interest Group on Data Communication · doi:10.1145/2785956.2787505 · Cited by 246 · cited 1×
“However, when moving a tremendous amount of data across the wide area network (WAN), both monetary cost and transmission delay can be prohibitively high, and the privacy leakage can also be a major concern [5].”
BackgroundExploiting Massive D2D Collaboration for Energy-Efficient Mobile Edge Computing
Chen, X., Pu, L., Gao, L., Wu, W., & Wu, D. (2017) · IEEE Wireless Communications · doi:10.1109/mwc.2017.1600321 · Cited by 239 · cited 1×
“1, here an edge node can be nearby end-device connectable by device-to-device (D2D) communications [7], a server attached to an access point (e.”
BackgroundSelf-citationPrivacy-preserving deep learning
Shokri, R., & Shmatikov, V. (2015) · 2015 53rd Annual Allerton Conference on Communication, Control, and Computing (Allerton) · doi:10.1109/allerton.2015.7447103 · Cited by 223
sharma (2018) · doi:10.1109/mic.2018.112102519 · Cited by 220
On-Demand Deep Model Compression for Mobile Devices: A Usage-Driven Model Selection Framework
Liu, S., Lin, Y., Zhou, Z., Nan, K., Liu, H., & Du, J. (2018) · Proceedings of the 16th Annual International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3210240.3210337 · Cited by 217 · cited 1×
“To this end, the proposed automatic optimization framework AdaDeep [75] systematically formulates the goals and constraints on accuracy, latency, storage, and energy into a unified optimization problem and leverages DRL to effectively find a good combination of compression techniques.”
MethodSelf-citationCommunication Compression for Decentralized Training
Tang, H., Gan, S., Zhang, C., Zhang, T., & Liu, J. (2018) · arXiv · doi:10.48550/arxiv.1803.06443 · Cited by 206 · cited 1×
“[56] develop a framework of compressed, decentralized training and propose two different algorithms, called extrapolation compression and difference compression, respectively.”
Bandwidth-Efficient Live Video Analytics for Drones Via Edge Computing
Wang, J., Feng, Z., Chen, Z., George, S., Bala, M., Pillai, P., Yang, S.-W., & Satyanarayanan, M. (2018) · 2018 IEEE/ACM Symposium on Edge Computing (SEC) · doi:10.1109/sec.2018.00019 · Cited by 204 · cited 1×
“[98] optimize for the first hop wireless bandwidth of DNN inference.”
Not Just Privacy: Improving Performance of Private Deep Learning in Mobile Cloud
Wang, J., Zhang, J., Bao, W., Zhu, X., Cao, B., & Yu, P. S. (2018) · Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3219819.3220106 · Cited by 188 · cited 5×
“As [59] referred, the design of DNN splitting in Arden is inspired by transfer learning.”
BackgroundMethodInfluentialSelf-citationNeurosurgeon: Collaborative Intelligence Between the Cloud and Mobile Edge
Kang, Y., Hauswald, J., Gao, C., Rovinski, A., Mudge, T., Mars, J., & Tang, L. (2017) · ACM SIGPLAN Notices · doi:10.1145/3093336.3037698 · Cited by 185
Big Data Privacy Preserving in Multi-Access Edge Computing for Heterogeneous Internet of Things
Du, M., Wang, K., Chen, Y., Wang, X., & Sun, Y. (2018) · IEEE Communications Magazine · doi:10.1109/mcom.2018.1701148 · Cited by 178
Low Latency Geo-distributed Data Analytics
Pu, Q., Ananthanarayanan, G., Bodík, P., Kandula, S., Akella, A., Bahl, P., & Stoica, I. G. (2015) · ACM SIGCOMM Computer Communication Review · doi:10.1145/2829988.2787505 · Cited by 177
A Berkeley View of Systems Challenges for AI
Stoica, I., Song, D., Popa, R. A., Patterson, D., Mahoney, M. W., Katz, R., Joseph, A. D., Jordan, M., Hellerstein, J. M., Gonzalez, J. E., et al. (2017) · arXiv · doi:10.48550/arxiv.1712.05855 · Cited by 171 · cited 1×
“In December 2017, in a white paper, “A Berkeley View of Systems Challenges for AI” [39] published by the University of California Berkeley, the cloud–edge AI system is envisioned as an important research direction to achieve the goal of mission-critical and personalized AI.”
BackgroundJALAD: Joint Accuracy-And Latency-Aware Deep Structure Decoupling for Edge-Cloud Execution
Li, H., Hu, C., Jiang, J., Wang, Z., Wen, Y., & Zhu, W. (2018) · 2018 IEEE 24th International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/padsw.2018.8645013 · Cited by 160 · cited 2×
“[81] H. Li, C. Hu, J. Jiang, Z. Wang, Y. Wen, and W. Zhu, “JALAD: Joint accuracy-and latency-aware deep structure decoupling for edge-cloud execution,” 2018, arXiv:1812.10027.”
MethodInfluentialZeng, L., Li, E., Zhou, Z., & Chen, X. (2019) · IEEE Network · doi:10.1109/mnet.001.1800506 · Cited by 159 · cited 2×
“Model Early-Exit • Partial DNNs model inference • Accuracy-aware [10], [15], [78], [87]–[91]”
BackgroundSelf-citationKo, J. H., Na, T., Amir, M. F., & Mukhopadhyay, S. (2018) · 2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) · doi:10.1109/avss.2018.8639121 · Cited by 153 · cited 1×
“[83] propose an edge-host partitioning method, which combines model partition with lossy feature encoding.”
MethodPipeDream: Fast and Efficient Pipeline Parallel DNN Training
Harlap, A., Narayanan, D., Phanishayee, A., Seshadri, V., Devanur, N., Ganger, G., & Gibbons, P. (2018) · arXiv · doi:10.48550/arxiv.1806.03377 · Cited by 153 · cited 2×
“[61] design PipeDream, a system that supports pipelined training and automatically determines how to systematically split a given model across the available computing nodes.”
BackgroundCachier: Edge-Caching for Recognition Applications
Drolia, U., Guo, K., Tan, J., Gandhi, R., & Narasimhan, P. (2017) · 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs.2017.94 · Cited by 148
Peer-to-peer Federated Learning on Graphs
Lalitha, A., Kilinc, O. C., Javidi, T., & Koushanfar, F. (2019) · arXiv · doi:10.48550/arxiv.1901.11173 · Cited by 146 · cited 1×
“[48] propose a Bayesian-based distributed algorithm, in which each device updates its belief by aggregating information from its one-hop neighbors to train a model that best fits the observations over the entire network.”
BackgroundIONN: Incremental Offloading of Neural Network Computations from Mobile Devices to Edge Servers
Jeong, H.-J., Lee, H.-J., Shin, C. H., & Moon, S.-M. (2018) · Proceedings of the ACM Symposium on Cloud Computing · doi:10.1145/3267809.3267828 · Cited by 143 · cited 4×
“IONN [82] propose an incremental offloading technique for EI application.”
MethodInfluentialBig/little deep neural network for ultra low power inference
Park, E., Kim, D., Kim, S., Kim, Y.-D., Kim, G., Yoon, S., & Yoo, S. (2015) · 2015 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS) · doi:10.1109/codesisss.2015.7331375 · Cited by 130 · cited 1×
“[102] propose a big/little DNN model selection framework.”
BackgroundMathur, A., Lane, N. D., Bhattacharya, S., Boran, A., Forlivesi, C., & Kawsar, F. (2017) · Proceedings of the 15th Annual International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3081333.3081359 · Cited by 128 · cited 4×
“DeepEye first segregates DNN layers of all tasks into two pools: convolution layers and fully connected layers.”
InfluentialAdaptive deep learning model selection on embedded systems
Taylor, B., Marco, V. S., Wolff, W., Elkhatib, Y., & Wang, Z. (2018) · ACM SIGPLAN Notices · doi:10.1145/3299710.3211336 · Cited by 119
MeDNN: A distributed mobile system with enhanced partition and deployment for large-scale DNNs
Mao, J., Yang, Z., Wen, W., Wu, C., Song, L., Nixon, K. W., Chen, X., Li, H., & Chen, Y. (2017) · 2017 IEEE/ACM International Conference on Computer-Aided Design (ICCAD) · doi:10.1109/iccad.2017.8203852 · Cited by 97
Self-citationTowards Real-Time Object Detection on Embedded Systems
Mao, H., Yao, S., Tang, T., Li, B., Yao, J., & Wang, Y. (2018) · IEEE Transactions on Emerging Topics in Computing · doi:10.1109/tetc.2016.2593643 · Cited by 94 · cited 2×
“Applying various convolutional filters, CNN models can capture the high-level representation of the input data, making it most popular for computer vision tasks, e.g., image classification (e.g., AlexNet [18], VGG network [19], ResNet [20], and MobileNet [21]) and object detection (e.g., Fast R-CNN [22], YOLO [23], an…”
On-Device Federated Learning via Blockchain and its Latency Analysis.
Kim, H., Park, J., Bennis, M., & Kim, S. (2018) · Cited by 94
eSGD: Communication Efficient Distributed Deep Learning on the Edge
Tao, Z., & Li, Q. (2018) · Cited by 92
FoggyCache: Cross-Device Approximate Computation Reuse
Guo, P., Hu, B., Li, R., & Hu, W. (2018) · Proceedings of the 24th Annual International Conference on Mobile Computing and Networking · doi:10.1145/3241539.3241557 · Cited by 91 · cited 5×
“There are two challenges in FoggyCache: one is that the input data distribution is unknown so the problem is how to index the input data with a constant lookup quality and the other is how to represent the similarity of the input data.”
MethodInfluentialAdaptive deep learning model selection on embedded systems
Taylor, B., Marco, V. S., Wolff, W., Elkhatib, Y., & Wang, Z. (2018) · Proceedings of the 19th ACM SIGPLAN/SIGBED International Conference on Languages, Compilers, and Tools for Embedded Systems · doi:10.1145/3211332.3211336 · Cited by 90 · cited 1×
“[103] point out that different DNN models (e.”
BackgroundTowards Wearable Cognitive Assistance
Ha, K., Chen, Z., Hu, W., et al. (2013) · doi:10.21236/ada591470 · Cited by 90
Li, Y., Park, J., Alian, M., Yuan, Y., Qu, Z., Pan, P., Wang, R., Schwing, A. G., Esmaeilzadeh, H., & Kim, N. S. (2018) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro.2018.00023 · Cited by 87 · cited 1×
“INCEPTIONN [68] Hybrid Level-5 • Maximizing the opportunities for”
BackgroundHow to scale distributed deep learning?
Jin, P. H., Yuan, Q., Iandola, F., & Keutzer, K. (2016) · arXiv · doi:10.48550/arxiv.1611.04581 · Cited by 85 · cited 25×
“Amiri and Gunduz [57] implement DSGD at the wireless edge with the help of a remote parameter server and further develop DSGD in digital and analog schemes, respectively.”
MethodInfluentialGossipGraD: Scalable Deep Learning using Gossip Communication based Asynchronous Gradient Descent
Daily, J., Vishnu, A., Siegel, C., Warfel, T., & Amatya, V. (2018) · arXiv · doi:10.48550/arxiv.1803.05880 · Cited by 80 · cited 2×
“[66] show that the trivial gossip-based algorithms at scale lead to a communication imbalance, poor convergence, and heavy communication overhead.”
BackgroundMethodGossip training for deep learning
Blot, M., Picard, D., Cord, M., & Thome, N. (2016) · arXiv · doi:10.48550/arxiv.1611.09726 · Cited by 77 · cited 1×
“Inspired by this, Gossip SGD (GoSGD) [64] is proposed to train DNN models in an asynchronous and decentralized way.”
MethodpCAMP: Performance Comparison of Machine Learning Packages on the Edges
Zhang, X., Wang, Y., & Shi, W. (2019) · arXiv · doi:10.48550/arxiv.1906.01878 · Cited by 77 · cited 1×
“, TensorFlow Lite, Caffe2, CoreML, and MXNet); however, empirical measurements [115] show that there is no single winner that can outperform other frameworks in all metrics.”
BackgroundMainstream: dynamic stem-sharing for multi-tenant video processing
Jiang, A. H., Wong, D., Canel, C., Tang, L., Misra, I., Kaminsky, M., Kozuch, M. A., Pillai, P. S., Andersen, D. G., & Ganger, G. R. (2018) · USENIX Annual Technical Conference · Cited by 73
Big/little deep neural network for ultra low power inference
Park, E., Kim, D., Kim, S., et al. (2015) · Seoul National University Open Repository (Seoul National University) · doi:10.5555/2830840.2830854 · Cited by 67
Designing adaptive neural networks for energy-constrained image classification
Stamoulis, D., Chin, T.-W. (., Prakash, A. K., Fang, H., Sajja, S., Bognar, M., & Marculescu, D. (2018) · Proceedings of the International Conference on Computer-Aided Design · doi:10.1145/3240765.3240796 · Cited by 63 · cited 1×
“[106] cast the adaptive DNN model selection issue as a hyperparameter optimization problem by taking into account the accuracy and communication constraints imposed by the devices.”
MethodThe cascading neural network: building the Internet of Smart Things
Leroux, S., Bohez, S., De Coninck, E., Verbelen, T., Vankeirsbilck, B., Simoens, P., & Dhoedt, B. (2017) · Knowledge and Information Systems · doi:10.1007/s10115-017-1029-1 · Cited by 59 · cited 1×
“For example, the cascading network [91] simply adds the MP layer and fully connected layer to the standard DNN model and achieves a speedup of 20%.”
BackgroundEdge Intelligence: On-Demand Deep Learning Model Co-Inference with Device-Edge SynergyPreprint
Li, E., Zhou, Z., & Chen, X. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.07840 · Cited by 59
Self-citationGeorgiev, P., Lane, N. D., Rachuri, K. K., & Mascolo, C. (2016) · Proceedings of the 22nd Annual International Conference on Mobile Computing and Networking · doi:10.1145/2973750.2973777 · Cited by 58 · cited 1×
“LEO [113] is a novel sensing algorithm scheduler that maximizes the performance for multiple continuous mobile sensor applications by partitioning the sensing algorithm execution and distributing tasks on CPU, coprocessor, GPU and the cloud.”
BackgroundOptimizing Grouped Aggregation in Geo-Distributed Streaming Analytics
Heintz, B., Chandra, A., & Sitaraman, R. K. (2015) · ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC) · doi:10.1145/2749246.2749276 · Cited by 53 · cited 1×
“Toward this goal, the conventional wisdom is to transport the data bulks from the IoT devices to the cloud datacenters for analytics [4].”
BackgroundEdgeOS_H: A Home Operating System for Internet of Everything
Cao, J., Xu, L., Abdallah, R., & Shi, W. (2017) · IEEE International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs.2017.325 · Cited by 51
Privacy-Preserving Machine Learning Based Data Analytics on Edge Devices
Zhao, J., Mortier, R., Crowcroft, J., & Wang, L. (2018) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3278721.3278778 · Cited by 50 · cited 1×
“Zoo [67] Hybrid Level-4 • Reducing communication cost • Privacy preservation • Federated Learning • Composable services • Processes each image within constant time despite the size difference of images”
BackgroundA Dynamic Deep Neural Network Design for Efficient Workload Allocation in Edge Computing
Lo, C., Su, Y.-Y., Lee, C.-Y., & Chang, S.-C. (2017) · 2017 IEEE International Conference on Computer Design (ICCD) · doi:10.1109/iccd.2017.49 · Cited by 48 · cited 1×
“[90] propose adding an authentic operation (AO) unit to the basic BranchyNet model.”
Data-driven Task Allocation for Multi-task Transfer Learning on the Edge
Chen, Q., Zheng, Z., Hu, C., Wang, D., & Liu, F. (2019) · IEEE International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs.2019.00107 · Cited by 46 · cited 1×
“[44] provide extensive studies on the performance (in both accuracy and convergence speed) of transfer learning, considering different student network architectures and different techniques for transferring knowledge from teacher to student.”
BackgroundLow-Latency Broadband Analog Aggregation for Federated Edge Learning.Preprint
Zhu, G., Wang, Y., & Huang, K. (2018) · arXiv (Cornell University) · Cited by 43
Are Existing Knowledge Transfer Techniques Effective for Deep Learning with Edge Devices?
Sharma, R., Biookaghazadeh, S., Li, B., & Zhao, M. X. (2018) · IEEE International Conference on Edge Computing and Communications (EDGE) · doi:10.1109/edge.2018.00013 · Cited by 41
EdgeOSH: A Home Operating System for Internet of Everything
Cao, J., Zhang, Q., & Shi, W. (2018) · SpringerBriefs in computer science · doi:10.1007/978-3-030-02083-5_2 · Cited by 34
Broadband Analog Aggregation for Low-Latency Federated Edge Learning (Extended Version)Preprint
Zhu, G., Wang, Y., & Huang, K. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1812.11494 · Cited by 33
(2017) · doi:10.1145/3132211.3134456 · Cited by 32
IF-CNN: Image-Aware Inference Framework for CNN With the Collaboration of Mobile Devices and Cloud
Shu, G., Liu, W., Zheng, X., & Li, J. (2018) · IEEE Access · doi:10.1109/access.2018.2880196 · Cited by 30 · cited 2×
“Similarly, IF-CNN [105] also trains a model selector called the recognition predictor (RP) to change the model used in the task.”
MethodChen, T. Y.-H., Balakrishnan, H., Ravindranath, L., & Bahl, P. (2016) · GetMobile Mobile Computing and Communications · doi:10.1145/2972413.2972423 · Cited by 28
Accelerating Deep Learning Workloads Through Efficient Multi-Model Execution
Narayanan, D., Santhanam, K., Phanishayee, A., & Zaharia, M. (2018) · Cited by 27
Kang, Y., Hauswald, J., Gao, C., Rovinski, A., Mudge, T., Mars, J., & Tang, L. (2017) · ACM SIGOPS Operating Systems Review · doi:10.1145/3093315.3037698 · Cited by 27
A portable, automatic data qantizer for deep neural networks
Oh, Y. H., Quan, Q., Kim, D.-Y., Kim, S., Heo, J., Jung, S., Jang, J.-Y., & Lee, J. W. (2018) · International Conference on Parallel Architectures and Compilation Techniques (PACT) · doi:10.1145/3243176.3243180 · Cited by 24
(2018) · Cited by 22
FFS-VA: A Fast Filtering System for Large-scale Video Analytics
Zhang, C., Cao, Q., Jiang, H., Zhang, W., Li, J., & Yao, J. (2018) · Proceedings of the 47th International Conference on Parallel Processing · doi:10.1145/3225058.3225103 · Cited by 16 · cited 3×
“Available: https://arxiv.org/abs/1811.01268 [100] C. Zhang, Q. Cao, H. Jiang, W. Zhang, J. Li, and J. Yao, “FFS-VA: A fast filtering system for large-scale video analytics,” in Proc.”
BackgroundInfluentialKLRA: A Kernel Level Resource Auditing Tool For IoT Operating System Security
Li, D., Zhang, Z., Liao, W., et al. (2018) · IEEE/ACM Symposium on Edge Computing (SEC) · doi:10.1109/sec.2018.00058 · Cited by 15 · cited 1×
“The open nature of edge computing imposes that the decentralized trust is required such that the EI services provided by different entities are trustworthy [121].”
BackgroundReXCam: Resource-Efficient, Cross-Camera Video Analytics at Enterprise Scale.Preprint
Jain, S., Jiang, J., Shu, Y., et al. (2018) · arXiv (Cornell University) · Cited by 12
ReXCam: Resource-Efficient, Cross-Camera Video Analytics at Scale
Jain, S., Zhang, X., Zhou, Y., Ananthanarayanan, G., Jiang, J., Shu, Y., & Gonzalez, J. (2018) · arXiv · doi:10.48550/arxiv.1811.01268 · Cited by 5
Mitigating Multi-tenant Interference in Continuous Mobile Offloading
Fang, Z., Luo, M., Yu, T., Mengshoel, O. J., Srivastava, M. B., & Gupta, R. K. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-94295-7_2 · Cited by 4 · cited 1×
“[109] Z.”
Mitigating multi-tenant interference on mobile offloading servers
Fang, Z., Luo, M., Yu, T., Mengshoel, O. J., Srivastava, M. B., & Gupta, R. K. (2017) · Symposium on Cloud Computing · doi:10.1145/3127479.3132563 · Cited by 2
In-Edge AI: Intelligentizing Mobile Edge Computing, Caching and Communication by Federated Learning
Wang, X., Han, Y., Wang, C., et al. (2018) · arXiv · doi:10.48550/arxiv.1809.07857
Machine Learning at the Wireless Edge: Distributed Stochastic Gradient Descent Over-the-Air
Amiri, M. M., & Gunduz, D. (2019) · arXiv · doi:10.48550/arxiv.1901.00844
A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics
Osia, S. A., Shamsabadi, A. S., Sajadmanesh, S., et al. (2017) · arXiv · doi:10.48550/arxiv.1703.02952
JALAD: Joint Accuracy- and Latency-Aware Deep Structure Decoupling for Edge-Cloud Execution
Li, H., Hu, C., Jiang, J., et al. (2018) · arXiv · doi:10.48550/arxiv.1812.10027
Ko, J. H., Na, T., Amir, M. F., et al. (2018) · arXiv · doi:10.48550/arxiv.1802.03835
Flagged first, then by citations
28 of the 168 references couldn’t be matched to a known paper.
Research works printed but not linked: 11
- 0, 3 AI Trends for Enterprise Computing
- 0, Microsoft Interactive Cloud Gaming
- 0, Democratizing AI
- hsieh, 2017, Gaia: Geo-distributed machine learning approaching LAN speeds, Proc NSDI, 629
- 2018, 5 Trends Emerge in the Gartner Hype Cycle for Emerging Technologies
- 0, Cisco global cloud index Forecast and methodology 2016–2021
- 0, Accelerating AI on the Intelligent Edge
- 2017, IEC White Paper Edge Intelligence
- 0, Edge Intelligence for Industrial Internet of Things
- han, 2015, Learning both weights and connections for efficient neural network, Proc Adv Neural Inf Process Syst, 1135
- 2018, 5 Trends Emerge in the Gartner Hype Cycle for Emerging Technologies
Cited by
Sorted by citations
Federated Learning in Mobile Edge Networks: A Comprehensive Survey
Lim, W. Y. B., Luong, N. C., Hoang, D. T., Jiao, Y., Liang, Y.-C., Yang, Q., Niyato, D., & Miao, C. (2020) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2020.2986024 · Cited by 2,565
“In particular, the authors in [14], [37], and [38] study the architectures and process of training and inference at edge networks without considering the challenges to FL implementation.”
BackgroundConvergence of Edge Computing and Deep Learning: A Comprehensive Survey
Wang, X., Han, Y., Leung, V. C. M., Niyato, D., Yan, X., & Chen, X. (2020) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2020.2970550 · Cited by 1,498
“Besides, discussions about DL inference and training are the main contribution of [19]–[21].”
BackgroundSelf-citation6G Internet of Things: A Comprehensive Survey
Nguyen, D. C., Ding, M., Pathirana, P. N., Seneviratne, A. P., Li, J., Niyato, D. T., Dobre, O. A., & Poor, H. V. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3103320 · Cited by 1,468
“6G is envisioned to be a unified communication platform, not only on the land but also on the sky (such as flight) and underwater (such as voyage), to achieve an extremely broad coverage and ubiquitous connectivity for fully supporting future IoT applications.”
BackgroundAcosta, J. N., Falcone, G. J., Rajpurkar, P., & Topol, E. J. (2022) · Nature Medicine · doi:10.1038/s41591-022-01981-2 · Cited by 1,365
“An additional layer of safety can be obtained by leveraging novel developments in edge computing 171 .”
BackgroundEdge Intelligence: The Confluence of Edge Computing and Artificial Intelligence
Deng, S., Zhao, H., Fang, W. J., Yin, J., Dustdar, S., & Zomaya, A. Y. (2020) · IEEE Internet of Things Journal · doi:10.1109/jiot.2020.2984887 · Cited by 1,280
“For example, Zhou et al. believe that the scope of Edge Intelligence should not be restricted to running AI models solely on the edge servers or devices but in the manner of the collaboration of edge and cloud [4].”
BackgroundMethod
Metadata
- Authors
- Zhou, Z., Chen, X., Li, E., Zeng, L., Luo, K., & Zhang, J.
- Affiliations
- Sun Yat-sen University (CN)
- Arizona State University (US)
- Venue
- Proceedings of the IEEE
- Published
- 2019-08
- Volume
- 107
- Issue
- 8
- Pages
- 1738-1762
- ISSN
- 0018-9219 (ISSN-L)
- 1558-2256
- OpenAlex
- W2950865323
- Semantic Scholar
- 928cd808aba140ec298508df87c5579811ff2f41
- Cited by
- 2,455
- References
- 168
- Citations by year
- 2026: 455 · 2025: 352 · 2024: 332 · 2023: 363 · 2022: 370 · 2021: 341 · 2020: 215 · 2019: 26
- FWCI
- 130.23
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Proceedings of the IEEE (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Basic Research Program of China (2017YFB1001703)
- National Natural Science Foundation of China (U1711265, 61802449)
- Program for Guangdong Introducing Innovative and Entrepreneurial Teams (2017ZT07X355)
- Guangdong Natural Science Funds (2018A030313032)
- Fundamental Research Funds for the Central Universities (17lgjc40)
- Army Research Office (W911NF-16-1-0448)
- Defense Threat Reduction Agency (HDTRA1-13-1-0029)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Information Systems › Internet of Things and AI
- Keywords
- edge AI, edge computing, edge computing architecture, deep learning training, deep learning inference, Internet of Things, mobile computing
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
