Enabling deep learning at the IoT edge
The DOI matches its Crossref record; 12 of the 13 references it cites were found in a registry or an index.
Abstract
Deep learning algorithms have demonstrated super-human capabilities in many cognitive tasks, such as image classification and speech recognition. As a result, there is an increasing interest in deploying neural networks (NNs) on low-power processors found in always-on systems, such as those based on Arm Cortex-M microcontrollers. In this paper, we discuss the challenges of deploying neural networks on microcontrollers with limited memory, compute resources and power budgets. We introduce CMSIS-NN, a library of optimized software kernels to enable deployment of NNs on Cortex-M cores. We also present techniques for NN algorithm exploration to develop light-weight models suitable for resource constrained systems, using keyword spotting as an example.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
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 1×
“Depthwise separable con volutional neural networks (DS-CNN), inspired from the efficient convolutions in Mobilenet architecture [8], decompose standard 3-”
MethodTensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., et al. (2016) · arXiv · doi:10.48550/arxiv.1603.04467 · Cited by 9,408
Edge 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
Fixed Point Quantization of Deep Convolutional Networks
Lin, D. D., Talathi, S. S., & Annapureddy, V. S. (2015) · arXiv · doi:10.48550/arxiv.1511.06393 · Cited by 684
Throughput-Optimized OpenCL-based FPGA Accelerator for Large-Scale Convolutional Neural Networks
Suda, N., Chandra, V., Dasika, G., Mohanty, A., Ma, Y., Vrudhula, S., Seo, J.-S., & Cao, Y. (2016) · Proceedings of the 2016 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/2847263.2847276 · Cited by 553
Self-citationSmall-footprint keyword spotting using deep neural networks
Chen, G., Parada, C., & Heigold, G. (2014) · 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2014.6854370 · Cited by 545
Convolutional neural networks for small-footprint keyword spotting
Sainath, T. N., & Parada, C. (2015) · Interspeech 2015 · doi:10.21437/interspeech.2015-352 · Cited by 521 · cited 6×
“Table 3 summarizes the accuracy, mem ory requirements and operations per inference for the network architectures for KWS from literature [ 6, 7, 13, 17].”
BackgroundMethodInfluentialHello Edge: Keyword Spotting on Microcontrollers
Zhang, Y., Suda, N., Lai, L., & Chandra, V. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.07128 · Cited by 334
Self-citationCMSIS-NN: Efficient Neural Network Kernels for Arm Cortex-M CPUs
Oliveira, A., Costa, D., Moreira, G., Martins, J., & Pinto, S. (2018) · arXiv (Cornell University) · doi:10.4230/oasics.ng-res.2025.4 · Cited by 288
Self-citationConvolutional Recurrent Neural Networks for Small-Footprint Keyword Spotting
Arık, S. Ö., Kliegl, M., Child, R., Hestness, J., Gibiansky, A., Fougner, C., Prenger, R., & Coates, A. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-1737 · Cited by 148
Research works printed but not linked: 4
- Liangzhen Lai, Naveen Suda, and Vikas Chandra. 2017. Deep Convolutional Neural Network Inference with Floating-point Weights and Fixed-point Activations. arXiv preprint arXiv:1703.03073 (2017). · arXiv
- 10.5555/3014904.3014977 · doi:10.5555/3014904.3014977
- 10.5555/3045390.3045690 · doi:10.5555/3045390.3045690
- Pete Warden. 2017. Speech Commands: A public dataset for single-word speech recognition. Dataset available from http://download.tensorflow.org/data/speech_commands_v0.01.tar.gz (2017).
Cited by
Deep Learning With Edge Computing: A Review
Chen, J., & Ran, X. (2019) · Proceedings of the IEEE · doi:10.1109/jproc.2019.2921977 · Cited by 1,540
“Lai and Suda [72] provided CMSISNN, a library for ARM Cortex-M processors that maximize DNN performance through quantization.”
BackgroundInfluentialConvergence 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
“A software framework is valuable if it 1) provides a library of optimized software kernels to enable deployment of DL [150]; 2) automatically compresses DL models into smaller dense matrices by finding the minimum number of non-redundant hidden elements [151]; 3) performs quantization and coding on”
MethodEfficient Acceleration of Deep Learning Inference on Resource-Constrained Edge Devices: A Review
Shuvo, M. M. H., Islam, S. K., Cheng, J., & Morshed, B. I. (2023) · Proceedings of the IEEE · doi:10.1109/jproc.2022.3226481 · Cited by 436
Machine Learning at the Network Edge: A Survey
Murshed, M. G. S., Murphy, C., Hou, D., Khan, N., Ananthanarayanan, G., & Hussain, F. (2022) · ACM Computing Surveys · doi:10.1145/3469029 · Cited by 402
“otics,andmedicalinstruments. LaiandSudadesigned asetofefficientneuralnetworkkernelscalledCMSIS-NN37 tomaximizetheperformanceofaneuralnetwork usinglimitedmemoryandcomputeresourcesofanARMCortex-Mprocessor[74]. 5.2.3 EmbeddedGPUs TheRaspberryPi,asingle-boardcomputerdevelopedbytheRaspberryPiFoundation,isoneofthemost commond…”
InfluentialQuantization and Deployment of Deep Neural Networks on Microcontrollers
Novac, P.-E., Boukli Hacene, G., Pegatoquet, A., Miramond, B., & Gripon, V. (2021) · Sensors · doi:10.3390/s21092984 · Cited by 210
“STM32Cube.AI also has an optimized inference engine that seems to be partially based on CMSIS-NN.”
MethodInfluentialSyed, N., Ge, M., & Baig, Z. A. (2023) · Computer Networks · doi:10.1016/j.comnet.2023.109662 · Cited by 105
CLIO: enabling automatic compilation of deep learning pipelines across IoT and cloud
Huang, J., Samplawski, C., Ganesan, D., Marlin, B. M., & Kwon, H. (2020) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3372224.3419215 · Cited by 99
“As a result, existing work on optimizing machine learning models for IoT has used simpler models and relied on hand-crafted optimizations rather state-of-art deep learning models and automated model compression techniques [26].”
BackgroundGraphHD: Efficient graph classification using hyperdimensional computing
Nunes, I., Heddes, M., Givargis, T., Nicolau, A. E., & Veidenbaum, A. V. (2022) · Design, Automation, and Test in Europe · doi:10.23919/date54114.2022.9774533 · Cited by 74
Blood Glucose Prediction in Type 1 Diabetes Using Deep Learning on the Edge
Zhu, T., Kuang, L., Li, K., Zeng, J., Herrero, P., & Georgiou, P. (2021) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas51556.2021.9401083 · Cited by 63
“Therefore, packing the front-end sensors together with the back-end deep learning algorithms [17] delivers an intuitive instruction on the edge inference with microcontroller units (MCUs), which stands out as a more resource-efficient solution within an IoT network.”
BackgroundReggiani, E., Pappalardo, A., Doblas, M., Moretó, M., Olivieri, M., Ünsal, O., & Cristal, A. (2023) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca56546.2023.10071076 · Cited by 54
“Application-specific libraries targeting commercial edge processors, such as Facebook QNNPack [23], Arm CMSIS-NN [37], and Google GEMMLowp [33] are often used to boost the performance of quantized DNNs on edge processors.”
MethodUCA-EHAR: A Dataset for Human Activity Recognition with Embedded AI on Smart Glasses
Novac, P.-E., Pégatoquet, A., Miramond, B., & Caquineau, C. (2022) · Applied Sciences · doi:10.3390/app12083849 · Cited by 43
“MicroAI can optionally make use of the CMSIS-NN [25] library for faster 8-or 16-bit fixed-point inference, taking advantage of the so-called DSP instructions available in the ARMv7E-M instruction set architecture of the Cortex-M4 core.”
MethodArtificial Intelligence for DC Arc Fault Detection in Photovoltaic Systems: A Comprehensive Review
Paul, K. C., Waldmann, D., Chen, C., Wang, Y., & Zhao, T. (2025) · IEEE Access · doi:10.1109/access.2025.3572521 · Cited by 32
“With decreasing memory and amount of parameters, the accuracy significantly drops for AI applications [130], [134].”
Analyzing Machine Learning on Mainstream Microcontrollers
Falbo, V., Apicella, T., Aurioso, D., Danese, L., Bellotti, F., Berta, R., & Gloria, A. D. (2020) · Lecture notes in electrical engineering · doi:10.1007/978-3-030-37277-4_12 · Cited by 30
Sensing beyond itself: Multi-functional use of ubiquitous signals towards wearable applications
Wang, Z., Li, J., Jin, Y., Wang, J., Yang, F. C., Li, G., Ni, X., & Ding, W. (2021) · Digital Signal Processing · doi:10.1016/j.dsp.2021.103091 · Cited by 27
Optimizing Convolutions for Deep Learning Inference on ARM Cortex-M Processors
Maciá-Lillo, A., Barrachina, S., Fabregat, G., & Dolz, M. F. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3395335 · Cited by 26
“3) optimized Convolutional Operators: Lai and Suda [24] presented CMSIS-NN, one of the first libraries for Cortex-M processors, which provides optimized kernels that maximise DNN performance while minimising memory requirements.”
MethodAIoT Solution Survey and Comparison in Machine Learning on Low-cost Microcontroller
Pham, H.-T., Nguyen, M.-A., & Sun, C. (2019) · International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) · doi:10.1109/ispacs48206.2019.8986357 · Cited by 24
Hyperdimensional computing: a framework for stochastic computation and symbolic AI
Heddes, M., Nunes, I., Givargis, T., Nicolau, A. E., & Veidenbaum, A. V. (2024) · Journal Of Big Data · doi:10.1186/s40537-024-01010-8 · Cited by 24
“Despite the functional accuracy achieved by GNNs, the high computational and energy cost of deep learning approaches make them difficult, or prohibitive, to be applied in real-world situations, such as those encountered in IoT and embedded applications [54, 63].”
BackgroundMonocular Depth Perception on Microcontrollers for Edge Applications
Peluso, V., Cipolletta, A., Calimera, A., Poggi, M., Tosi, F., Aleotti, F., & Mattoccia, S. (2021) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2021.3077395 · Cited by 23
“As a rule of thumb, accurate strategies, i.e. linear/asymmetric and non-linear quantization, achieve higher fidelity at the cost of lower performance [49].”
BackgroundMethodInfluentialFrom Zombies to Smart Devices: The Evolution of Dye-Sensitized Solar Cells for IoT Applications
Sasitharan, K., & Freitag, M. (2025) · ACS Applied Energy Materials · doi:10.1021/acsaem.5c00624 · Cited by 19
A Reinforcement Learning-Based Framework for the Exploitation of Multiple RATs in the IoT
Sandoval, R. M., Canovas-Carrasco, S., García-Sánchez, A.-J., & García-Haro, J. (2019) · IEEE Access · doi:10.1109/access.2019.2938084 · Cited by 19
“This is a process that takes much more time to run, and therefore, consumes much more energy (when compared to running it in an FPU).”
MethodBoosting Distributed Machine Learning Training Through Loss-tolerant Transmission Protocol
Chen, Z., Shi, L., Liu, X., Ai, X., Liu, S., & Xu, Y. (2023) · International Workshop on Quality of Service · doi:10.1109/iwqos57198.2023.10188699 · Cited by 17
“To improve DML communication, several works have been proposed such as parameter quantization [19], parameter pruning [20], gradient compression [21], and gradient quantization [22].”
MethodEnergy-Efficient Adaptive Machine Learning on IoT End-Nodes With Class-Dependent Confidence
Daghero, F., Burrello, A., Pagliari, D. J., Benini, L., Macii, E., & Poncino, M. (2020) · IEEE International Conference on Electronics, Circuits and Systems (ICECS) · doi:10.1109/icecs49266.2020.9294863 · Cited by 16
“Running Machine Learning (ML) inference directly on IoT edge devices can yield benefits in faster response times, improved data privacy, and higher energy efficiency, by avoiding the transmission of raw sensor data through energyhungry wireless channels [1], [2].”
BackgroundMethodRe-thinking computation offload for efficient inference on IoT devices with duty-cycled radios
Huang, J., Guan, H., & Ganesan, D. (2023) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3570361.3592514 · Cited by 15
“Recent years have seen substantial interest in deployingDeep Neural Network (DNN) models on resource-constrained Internet of Things (IoT) devices [24, 30, 33, 44].”
BackgroundWang, J., & Li, S. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2208.02765 · Cited by 15
“be incompatible with other accelerating methods for microcontrollers [3].”
BackgroundInfluentialCACTUS: Dynamically Switchable Context-aware micro-Classifiers for Efficient IoT Inference
Rastikerdar, M. M., Huang, J., Fang, S., Guan, H., & Ganesan, D. (2024) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3643832.3661888 · Cited by 14
SAVector: Vectored Systolic Arrays
Choi, S., Park, S., Park, J., Kim, J., Koo, G., Hong, S., Yoon, M. K., & Oh, Y. (2024) · IEEE Access · doi:10.1109/access.2024.3380433 · Cited by 13
“SAVector does not explicitly concatenate rows but instead uses a Height-Width-Channel (HWC) data format, a concept demonstrated in previous work [28], [29].”
MethodSen, T., & Shen, H. (2022) · IEEE International Conference on Distributed Computing Systems (ICDCS) · doi:10.1109/icdcs54860.2022.00062 · Cited by 12
Space-Efficient TREC for Enabling Deep Learning on Microcontrollers
Liu, J., Zhang, F., Guan, J., Sung, H.-H., Guo, X., Du, X., & Shen, X. (2023) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3582016.3582062 · Cited by 11
“There are some prior efforts trying to enable DNNs on microcontrollers, by creating light-weight neural networks [31, 60] and other designs [10, 14, 46].”
BackgroundUse of edge resources for DNN model maintenance in 5G IoT networks
Sung, J.-W., & Han, S.-J. (2024) · Cluster Computing · doi:10.1007/s10586-023-04236-y · Cited by 11
“In [34] [38], various techniques for shrinking large DNN models are discussed.”
Integrating Embedded Neural Networks and Self-Mixing Interferometry for Smart Sensors Design
Novac, P.-E., Rodriguez, L., & Barland, S. (2024) · IEEE Sensors Applications Symposium (SAS) · doi:10.1109/sas60918.2024.10636497 · Cited by 9
“In order to speed the inference up, the CMSIS-NN [23] library is used.”
MethodNAIR: An Efficient Distributed Deep Learning Architecture for Resource Constrained IoT System
Xiao, Y., Zhang, D., Wang, Y., Dai, X., Huang, Z., Zhang, W., Yang, Y., Anjum, A., & Qin, F. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3375924 · Cited by 8
“Although many commercial off the shelf (COTS) devices [10] and published works [11], [12] have already demonstrated this possibility, the computing and memory resources of these devices are usually several orders lower than the classical deep learning platforms.”
BackgroundEdge of Innovation: Exploring the Potential of Edge Computing and Content Delivery Networks
Dash, Y., Kumar, N., & Abraham, A. (2024) · IEEE International Conference on Computing, Power and Communication Technologies (IC2PCT) · doi:10.1109/ic2pct60090.2024.10486411 · Cited by 8
Li, Q., Lin, S., Liu, C., Liu, Y., Qiao, F., Wang, Y., & Yang, H. (2020) · International Symposium on Low Power Electronics and Design (ISLPED) · doi:10.1145/3370748.3407001 · Cited by 8
“These NN-based models achieve significant latency and accuracy improvement over the conventional hidden markov model [19] [9] [21].”
Inverse-GMM: A Latency Distribution Shaping Method for Industrial Cooperative Deep Learning Systems
Qin, F., Xiao, Y., Sun, X., Dai, X., Zhang, W., & Shen, F. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3229448 · Cited by 8
“Although many Commercial off the Shelf (COTS) devices [2] and published work [3], [4] have demonstrated such possibility, the computation resources of these devices are several orders lower than the classical deep learning platforms.”
BackgroundMethodIEEE 802.11ah for Internet of Vehicles: Design Issues and Challenges
Pavan, B. S., Mahesh, M., & Harigovindan, V. P. (2020) · Unmanned system technologies · doi:10.1007/978-3-030-46335-9_4 · Cited by 8
Lakshaga Jyothi, M. (2021) · Türk bilgisayar ve matematik eğitimi dergisi · doi:10.17762/turcomat.v12i2.818 · Cited by 7
“Also in [69], the author tries to conduct experiments on using machine learning algorithm inferences on the edge in IoT devices for 3 algorithms and 10 different datasets on Raspberry pi.”
MethodEnergy-Accuracy Scalable Deep Convolutional Neural Networks: A Pareto Analysis
Peluso, V., & Calimera, A. (2019) · IFIP advances in information and communication technology · doi:10.1007/978-3-030-23425-6_6 · Cited by 6
“This implementation applies for any general purposes MCUs (e.g. [ 21]) or application-specific processors with a multi-precision instruction-set (e.g. Google TPU [22]); it can also be extended to dedicated architecture (both ASIC or FPGA accelerators [23]).”
AI Workload Allocation Methods for Edge-Cloud Computing: A Review
Rafea, S. A., & Jasim, A. D. (2023) · Al-Iraqia Journal of Scientific Engineering Research · doi:10.58564/ijser.2.4.2023.125 · Cited by 6
Li, B., Zuo, K., Li, L., & Wu, Y. (2026) · Measurement Science and Technology · doi:10.1088/1361-6501/ae6469 · Cited by 3
“This process converts high-level deep learning models into low-level C code tailored for microcontroller execution, ensuring real-time performance under severe resource limitations [33].”
Low-Cost Image Search System on Off-Line Situation
Diana, M., Chikama, J., Amagasaki, M., & Iida, M. (2020) · Electronics · doi:10.3390/electronics9010153 · Cited by 3
“As a result of this condition, a deep learning model should fit with the limitation of memory on the microcontroller and limitation of resource in the edge device for computing the task such as real-time classification processing [11].”
BackgroundTensor Layout Optimization of Convolution for Inference on Digital Signal Processor
Zhang, X., Xiao, J., Zhang, X., Hu, Z., Zhu, H., Tian, Z., & Tan, G. (2019) · IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA) · doi:10.1109/ispa-bdcloud-sustaincom-socialcom48970.2019.00036 · Cited by 3
“It can process various kinds of algorithms with the complex digital signal in real-time [2], [23].”
BackgroundMicroAI : embedded artificial intelligence for human activity recognition on smart glasses
Novac, P.-E. (2022) · theses.fr (ABES) · doi:10.70675/77c14a3ez0f87z4c4dza876z9064b1b107ec · Cited by 3
Multimodal Deep Learning with Boosted Trees for Edge Inference
Chong, P., Wynter, L., & Chaudhury, B. (2023) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw60847.2023.00021 · Cited by 2
Microcontroller-Driven Voice-Based Smart Wheelchair Navigation with Neural Networks
Gambhir, V., Bisht, C. S., Goe, A., & Gupta, A. K. (2024) · International Conference on Computing, Communication and Networking Technologies (ICCCNT) · doi:10.1109/icccnt61001.2024.10725700 · Cited by 2
“One such application is KWS on microcontrollers, which has seen remarkable progress due to the integration of neural networks (NNs) on low-power processors [1], [2].”
BackgroundWalinjkar, A. (2019) · Open Access at Essex (University of Essex) · Cited by 2
Otomatisasi Penjadwalan Pemberian Pakan Ikan Menggunakan Fuzzy Logic Rules Based System
Hantoro, K. (2024) · Journal of Informatics and Information Security · doi:10.31599/f1h59d54 · Cited by 1
Machine Learning Frameworks and Algorithms for Fog and Edge Computing
Perumalla, M. M. R., Singh, S. K., Khamparia, A., Goyal, A., & Mishra, A. (2020) · Wiley eBooks · doi:10.1002/9781119670087.ch4 · Cited by 1
Image Search System Based on Feature Vectors of Convolutional Neural Network
Diana, M., Amagasaki, M., & Iida, M. (2020) · IEEE Region 10 Conference (TENCON) · doi:10.1109/tencon50793.2020.9293773 · Cited by 1
“The edge device offering the realtime application, it is a chance to build a real-time classification [6].”
BackgroundA Distributed Architecture for Cooperative Deep Learning System in Intelligent Vehicle Systems
Xiao, Y., Qin, F., Sun, X., & Wang, Y. (2022) · IEEE International Conference on Networking, Architecture and Storage (NAS) · doi:10.1109/nas55553.2022.9925525 · Cited by 1
“Although many Commercial off the Shelf (COTS) devices [3] and published work [4] [5] have demonstrated such possibility, the computation resources”
BackgroundA Distributed Deep Learning System With Controlled Intermediate Representation
Xiao, Y., Wang, Y., Huang, Z., Shen, F., & Qin, F. (2023) · IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10448716 · Cited by 1
“Although many Commercial off the Shelf (COTS) devices [10] and published work [11] [12] have demonstrated the possibility, the computing and memory resources of these devices are several orders lower than classical deep learning platform.”
BackgroundRuntime Adaptation in Embedded Computing Systems using Markov Decision Processes
Sapio, A. E. (2019) · University Libraries (University of Maryland) · doi:10.13016/6bqg-pdxr
Artificial Intelligence Inference in Edge
Wang, X., Han, Y., Leung, V. C. M., Niyato, D. T., Yan, X., & Chen, X. (2020) · Springer eBooks · doi:10.1007/978-981-15-6186-3_5
Platform-specific model compression for deep neural networks with joint methods
Lin, S. (2020) · Digital Repository Service (Northeastern University) · doi:10.17760/d20398267
Optimization Tools for ConvNets on the Edge
Peluso, V., Macii, E., & Calimera, A. (2020) · IFIP/IEEE International Conference on Very Large Scale Integration · doi:10.1109/vlsi-soc46417.2020.9344075
Wang, Z., Li, J., Jin, Y., Wang, J., Yang, F. C., Li, G., Ni, X., & Ding, W. (2022) · Digital Signal Processing · doi:10.1016/j.dsp.2022.103571
Image Processing on Resource-Constrained Devices
Tolia, D., Jagadeeshwar, S., Kumar, J., Arvind, P., & Yadav, A. R. (2023) · BENTHAM SCIENCE PUBLISHERS eBooks · doi:10.2174/9789815080537123010017
Machine Learning on Standard Embedded Device
Selvam, U., Subramanian, P. M., & Rajeswari, A. (2023) · International Journal of Computer Applications · doi:10.5120/ijca2023922911
Efficient hardware acceleration of deep neural networks via arithmetic complexity reduction
Reggiani, E. (2023) · UPCommons institutional repository (Universitat Politècnica de Catalunya) · doi:10.5821/dissertation-2117-404668
Hanreich, M., Krauss, O., & Zwettler, G. A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-82957-4_15
Region-based Convolutional Neural Network and Implementation of the Network Through Zedboard Zynq
ISLAM, M. M. (2019) · Purdue e-Pubs (Purdue University System) · doi:10.25394/pgs.7732307
Energy Analysis of Row-Stationary Dataflow in CNN Computation on Microcontrollers
Koppala, M. R. (2024) · Qucosa - Monarch (Chemnitz University of Technology) · doi:10.60687/2025-0048
Efficient and Accurate Graph Classification with Hyperdimensional Computing on FPGA
Arockiaraj, J., Parikh, D., & Prasanna, V. K. (2026) · ACM International Conference on Computing Frontiers (CF) · doi:10.1145/3801487.3801824
Real-time smart self-mixing interferometry sensor with embedded neural network
Novac, P.-E., Rodriguez, L., & Barland, S. (2026) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2026.3718110
“In order to speed the inference up, we use the CMSIS-NN 1 [28] library.”
Vertical Optimizations of Convolutional Neural Networks for Embedded Systems
Cipolletta, A. (2022) · PORTO@Iris
Metadata
- Authors
- Lai, L., & Suda, N.
- Affiliations
- Arm Inc.
- ORCID
- Lai, Liangzhen
- Venue
- Proceedings of the International Conference on Computer-Aided Design
- Publisher
- ACM
- Published
- 2018-11-05
- Pages
- 1-6
- OpenAlex
- W2900071409
- Cited by
- 49
- References
- 13
- Citations by year
- 2026: 2 · 2025: 3 · 2024: 9 · 2023: 10 · 2022: 5 · 2021: 4 · 2020: 12 · 2019: 4
- FWCI
- 1.51
- Type
- conference paper
- Language
- en
- Open access
- closed
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Keywords
- deep learning deployment, edge computing, Internet of Things, neural network inference, resource-constrained systems, keyword spotting, lightweight neural networks, low-power processors
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.
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.