Embedded Intelligence on FPGA: Survey, Applications and Challenges
Open accessThe DOI matches its Crossref record; 90 of the 93 references it cites were found in a registry or an index; 36 of the 140 sources its claims cite were checked for support.
Abstract
Embedded intelligence (EI) is an emerging research field and has the objective to incorporate machine learning algorithms and intelligent decision-making capabilities into mobile and embedded devices or systems. There are several challenges to be addressed to realize efficient EI implementations in hardware such as the need for: (1) high computational processing; (2) low power consumption (or high energy efficiency); and (3) scalability to accommodate different network sizes and topologies. In recent years, an emerging hardware technology which has demonstrated strong potential and capabilities for EI implementations is the FPGA (field programmable gate array) technology. This paper presents an overview and review of embedded intelligence on FPGA with a focus on applications, platforms and challenges. There are four main classification and thematic descriptors which are reviewed and discussed in this paper for EI: (1) EI techniques including machine learning and neural networks, deep learning, expert systems, fuzzy intelligence, swarm intelligence, self-organizing map (SOM) and extreme learning; (2) applications for EI including object detection and recognition, indoor localization and surveillance monitoring, and other EI applications; (3) hardware and platforms for EI; and (4) challenges for EI. The paper aims to introduce interested researchers to this area and motivate the development of practical FPGA solutions for EI deployment.
Overview
CoChat Summary
TL;DR
This review surveys embedded intelligence (EI) on FPGA across techniques, applications, hardware platforms, and challenges, arguing that FPGA platforms are well suited for EI due to parallelism, flexibility, and energy efficiency. It synthesizes many cited accelerator studies and reports speed, power, and accuracy comparisons against CPU and GPU implementations.
Key findings
- FPGA is presented as a strong EI hardware candidate because it offers parallel computation, lower power consumption, short turnaround time versus ASIC, and reconfigurability; one comparison states FPGA can reach 10 times faster than CPU platforms, with the lowest level having same performance as GPU, and another states FPGA is 6.5 times more energy efficient than GPU.
- Cited accelerator results include CoNNA speeding up at least 3.08 times and up to 14.1 times faster than the MIT Eyeriss accelerator; an Arria 10 FPGA inference implementation being 6 times faster than GPU and 44 times faster than CPU; and a YOLO V2 Arria 10 design achieving 3.3 times more power efficiency than Titan X GPU and 418 times more power efficiency than CPU.
- Application-specific results include facial expression recognition achieving about 60.3% recognition efficiency at 200 MHz for 400 faces/second; a smart chair sitting posture ANN achieving 97.78% floating-point accuracy and 97.43% fixed-point accuracy with 7.35 mW dynamic power; and a millimeter-wave radar CNN acceleration being 30.42% faster than NVIDIA GPU with a 0.27% accuracy reduction to 87.81%.
- The paper identifies major EI challenges as high computational processing, low power consumption/high energy efficiency, and scalability, plus FPGA-specific issues such as reconfiguration time, programming difficulty, multiple-FPGA communication, limited on-chip memory, lack of built-in cache, and bandwidth typically two generations older than high-end GPU.
Methods
- Design: narrative survey/review of embedded intelligence on FPGA, organized around four classification and thematic descriptors: EI techniques, EI applications, EI hardware and platforms, and EI challenges.
- Data/sample: reviewed works and references spanning EI techniques such as machine learning and neural networks, deep learning, expert systems, fuzzy intelligence, swarm intelligence, SOM and extreme learning; applications including object detection/recognition, indoor localization and surveillance monitoring; and FPGA platforms from Xilinx, Intel/Altera, and Achronix.
- Main analysis: qualitative synthesis and tabular summaries, including Table 1 for classification/thematic descriptors and Table 2 for FPGA hardware platforms; no new empirical experiments or primary dataset are reported.
Limitations
- The paper does not state limitations of the review itself.
- Evident from the text: it is a narrative survey without a stated systematic search protocol, inclusion/exclusion criteria, or quantitative meta-analysis; reported speed, power, and accuracy figures come from heterogeneous cited studies and platforms, so they are not directly comparable.
- The paper discusses FPGA challenges such as reconfiguration time, programming cost, on-chip memory limits, and bandwidth constraints, but these are framed as challenges for FPGA-based EI rather than limitations of the review.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
28 Valid · 32 Not valid · 97 Not checked · 43 Own results and interpretations
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“To have high accuracy and efficiency of the LBS for indoor environments, fingerprint-based indoor localization is introduced by using scalable deep neural network (DNN) architecture implement together with denoising auto-encoder (SDNNLoc) based on FPGA [68].”
What the sources say
- Disputedcites [68] Xie, T., Jiang, H., Zhao, X., and Zhang, C. (2019). A Wi-Fi-Based wireless indoor position sensing system with multipath interference mitigation. Sensors, 19. sourceExperimental — accuracy not yet measured
The Introduction identifies the paper’s proposed system as RSSI-assisted time-difference-of-arrival (TDoA) positioning, and the Conclusions describe an improved cross-correlation method for mitigating multipath interference. The paper discusses fingerprinting and neural networks as other approaches, but it does not introduce the claimed FPGA-based SDNNLoc architecture with a denoising autoencoder. (LLM, conf=0.88)
Checked by a reasoning model — may be wrong.
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
Every source the paper cites for this claim was checked and supports it.
In the paper
“The result of implementing RCCM and AddNet was the ability to save resources by up to 50% more than a standard 8-bit quantized network with similar accuracy [34].”
What the sources say
- Supportedcites [34] Faraone, 2019, AddNet: Deep Neural Networks Using FPGA-Optimized Multipliers, IEEE Trans. Very Large Scale Integr. Syst. source
The paper’s opening summary (p. 1) reports up to 50% resource savings over traditional 8-bit quantized networks. It also says RCCM implementations other than the one with the lowest resource requirements achieve at least similar accuracy to an 8-bit design. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Both forward and backward propagation on a homogenous FPGA system were able to improve the training by using a DarkFPGA accelerator with batch level parallelism.”
What the sources say
- Supportedcites [32] Luo, 2020, Towards efficient deep neural network training by FPGA-based batch-level parallelism, J. Semicond. source
The Conclusion states that DarkFPGA uses batch-level parallelism to accelerate both forward and backward propagation on a homogeneous FPGA system. Section 8.2 also reports improved training performance from performing the entire process on the FPGA. (LLM, conf=0.94)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This approach gives better accuracy than the network that constrain weight parameters to have binary value.”
What the sources say
- Partly supportedcites [34] Faraone, 2019, AddNet: Deep Neural Networks Using FPGA-Optimized Multipliers, IEEE Trans. Very Large Scale Integr. Syst. sourceExperimental — accuracy not yet measured
The Accuracy Versus Area section (p. 12) reports that AddNet improves accuracy over 4-bit and 6-bit fixed-point implementations. It does not report an accuracy comparison with a network whose weights are constrained to binary values, so the specific comparator in the claim is unsupported. (LLM, conf=0.82)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“No overfitting phenomenon is observed for the model, as determined by the dropout that followed the layers in the neural network.”
What the sources say
- Supportedcites [59] 2019, FPGA platform applied for facial expression recognition system using convolutional neural networks, Procedia Comput. Sci. source
Section 4.1 (Testing) explicitly says the model did not exhibit overfitting and attributes this to dropout following the neural-network layers. The claim closely paraphrases that passage. (LLM, conf=0.95)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
51 more claims · 4 with their cited sources checked
Create a free account to see all 200 claims and to verify the 18 references still open.
- Probably not a substantial claim. 91
- It may state two separate claims as one. 81
- The quote only partly states this claim. 1
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
“Compared to 32-bit floating-point, a framework of 16-bit fixed-point multiplication reduces the power-consuming by 6.2 times, but with a reduction in accuracy of 0.5%.”
From Face Recognition Based on Deep Learning and FPGA for Ethnicity Identification (2022) · cites this paper as [43]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“As compared to the standard network, a reduction in parameters by 11.2 times was achieved using this fire module technique with no impact on the accuracy of classification [43].”
From Face Recognition Based on Deep Learning and FPGA for Ethnicity Identification (2022) · cites this paper as [43]
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.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 6. Challenges
“For example, a GPU Titan X uses GDDR5 but Arria 10 only uses DDR3.”
§ 5. FPGA Hardware and Platforms
“| Arria 10 | GT1150 | Intel | 1150 | 1518 | DDR4 |”
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 1×
“In the paper, it is shown that by running at the clock frequency of 125MHz it was able to become seven times more energy efficient than the LSTM reference for similar performance [81].”
BackgroundFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472 · cited 1×
“The algorithm applies anchor boxes to change object detection with different aspect ratios, and go through numerous pooling layers to acquire the classifier image [64].”
MethodMobileNetV2: Inverted Residuals and Linear Bottlenecks
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00474 · Cited by 28,755 · cited 2×
“The VGG-16 model per-trained is applied on MobileNet architecture to extract the feature maps.”
MethodParticle swarm optimization: An overview
Poli, R., Kennedy, J., & Blackwell, T. (2007) · Swarm Intelligence · doi:10.1007/s11721-007-0002-0 · Cited by 21,101
Xception: Deep Learning with Depthwise Separable Convolutions
Chollet, F. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.195 · Cited by 20,380 · cited 1×
“In the research [24], a depthwise separable convolution neural network on FPGA is being reviewed due to its ability to reduce the number of model parameters and improve calculation speeds, which makes it widely used in mobile edge computing applications such as Xception [25] and MobileNet [26].”
MethodEyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2016.2616357 · Cited by 3,247 · cited 2×
“Based on the analysis of the research, CoNNA is able to speed up at least 3.08 times and up to 14.1 times faster than the MIT Eyeriss Accelerator [20].”
BackgroundInfluentialAnt colony optimization theory: A survey
Dorigo, M., & Blum, C. (2005) · Theoretical Computer Science · doi:10.1016/j.tcs.2005.05.020 · Cited by 2,295 · cited 1×
“For the ACO method, a graph is used to represent the solution space [49].”
MethodMobileNetV2: Inverted Residuals and Linear BottlenecksPreprint
Sandler, M., Howard, A., Zhu, M., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1801.04381 · Cited by 2,255 · cited 2×
“The VGG-16 model per-trained is applied on MobileNet architecture to extract the feature maps.”
MethodThiNet: A Filter Level Pruning Method for Deep Neural Network Compression
Luo, J.-H., Wu, J., & Lin, W. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.541 · Cited by 1,832 · cited 1×
“The coarse-grained reconfigurable architecture is applied in CoNNA, which allows immediate selection of the CNN network [19].”
MethodThe Horus WLAN location determination system
Youssef, M., & Agrawala, A. (2005) · Proceedings of the 3rd international conference on Mobile systems, applications, and services · doi:10.1145/1067170.1067193 · Cited by 1,653 · cited 2×
“The performance evaluation is focused on the runtime of the SDNNLoc and compared with two other baseline systems, which are Horus [71] and DeepFi [72].”
MethodFINN: A Framework for Fast, Scalable Binarized Neural Network Inference
Umuroglu, Y., Fraser, N. J., Gambardella, G., Blott, M., Leong, P., Jahre, M., & Vissers, K. (2017) · Proceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/3020078.3021744 · Cited by 1,189
Object Recognition with Gradient-Based Learning
LeCun, Y., Haffner, P., Bottou, L., & Bengio, Y. (1999) · Lecture Notes in Computer Science · doi:10.1007/3-540-46805-6_19 · Cited by 947 · cited 1×
“SqueezeNet and LetNet are some of the examples using condensed parameter requirement techniques in their hardware accelerator design [27].”
MethodEdge Computing for Autonomous Driving: Opportunities and Challenges
Liu, S., Liu, L., Tang, J., Yu, B., Wang, Y., & Shi, W. (2019) · Proceedings of the IEEE · doi:10.1109/jproc.2019.2915983 · Cited by 778
Structured Pruning of Deep Convolutional Neural Networks
Anwar, S., Hwang, K., & Sung, W. (2017) · ACM Journal on Emerging Technologies in Computing Systems · doi:10.1145/3005348 · Cited by 698
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
New inspirations in swarm intelligence: a survey
Parpinelli, R. S., & Lopes, H. S. (2011) · International Journal of Bio-Inspired Computation · doi:10.1504/ijbic.2011.038700 · Cited by 402 · cited 1×
“In recent years, machine learning [1] and swarm intelligence [2] algorithms have demonstrated successes in many fields and solved practical problems in industry.”
BackgroundDeepFi: Deep learning for indoor fingerprinting using channel state information
Wang, X., Gao, L., Mao, S., & Pandey, S. (2015) · 2015 IEEE Wireless Communications and Networking Conference (WCNC) · doi:10.1109/wcnc.2015.7127718 · Cited by 381 · cited 2×
“The running time per location for the SDNNLoc is almost the same as the DeepFi system, where both use the DNN based technique.”
MethodSupport Vector Machine Solvers
Bottou, L., & Lin, C.-J. (2007) · Large-Scale Kernel Machines · doi:10.7551/mitpress/7496.003.0003 · Cited by 336
Training and Inference with Integers in Deep Neural Networks
Wu, S., Li, G., Chen, F., & Shi, L. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.04680 · Cited by 306
Caffeine: Toward Uniformed Representation and Acceleration for Deep Convolutional Neural Networks
Zhang, C., Sun, G., Fang, Z., Zhou, P., Pan, P., & Cong, J. (2019) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2017.2785257 · Cited by 270
Enhanced fingerprinting and trajectory prediction for IoT localization in smart buildings
Lin, K., Chen, M., Deng, J., Hassan, M. M., & Fortino, G. (2016) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2016.2543242 · Cited by 246 · cited 1×
“Indoor localization is challenging due to radio propagation environments such as shadowing, multipath effect, delay distortion and fading [67].”
BackgroundHardware accelerated convolutional neural networks for synthetic vision systems
Farabet, C., Martini, B., Akselrod, P., Talay, S., LeCun, Y., & Culurciello, E. (2010) · Proceedings of 2010 IEEE International Symposium on Circuits and Systems · doi:10.1109/iscas.2010.5537908 · Cited by 231 · cited 1×
“Due to consideration of reasonable compromise of precision and hardware efficiency [76], the feature maps and weight parameters are using 16-bit fixed-point number representation.”
MethodAn Iterative Procedure for Computing the Minimum of a Quadratic Form on a Convex Set
Gilbert, E. G. (1966) · SIAM Journal on Control · doi:10.1137/0304007 · Cited by 230 · cited 1×
“The Gilbert algorithm is used to solve the nearest point problem where all the computations are computed under the processing element (PE) [11].”
MethodScalable and modularized RTL compilation of Convolutional Neural Networks onto FPGA
Ma, Y., Suda, N., Cao, Y., Seo, J.-S., & Vrudhula, S. (2016) · 2016 26th International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl.2016.7577356 · Cited by 161
fpgaConvNet: Mapping regular and irregular convolutional neural networks on FPGAs
Venieris, S. I., & Bouganis, C.-S. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2844093 · Cited by 161
Kumar, R. (2017) · Machine Learning and Cognition in Enterprises · doi:10.1007/978-1-4842-3069-5_4 · Cited by 155
Doppler signatures of radar backscattering from objects with micro motions
Chen, V. C. (2008) · IET Signal Processing · doi:10.1049/iet-spr:20070137 · Cited by 146
A Lightweight YOLOv2: A Binarized CNN with A Parallel Support Vector Regression for an FPGA
Nakahara, H., Yonekawa, H., Fujii, T., & Sato, S. (2018) · Proceedings of the 2018 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/3174243.3174266 · Cited by 139 · cited 1×
“Titan X GPU is a reference platform for comparison and the framework is YOLO V2 [84].”
MethodWiGEM: a learning-based approach for indoor localization
Goswami, A., Ortiz, L. E., & Das, S. R. (2011) · Proceedings of the Seventh COnference on emerging Networking EXperiments and Technologies · doi:10.1145/2079296.2079299 · Cited by 126 · cited 1×
“The first stage is the offline database-building phase where an interested location area is defined for the collection of offline RSS fingerprints and building of baseline fingerprint database [69].”
BackgroundHardware Acceleration of Short Read Mapping
Olson, C. B., Kim, M., Clauson, C., Kogon, B., Ebeling, C., Hauck, S., & Ruzzo, W. L. (2012) · IEEE International Symposium on Field-Programmable Custom Computing Machines (FCCM) · doi:10.1109/fccm.2012.36 · Cited by 126 · cited 1×
“The temporary data will be stored as off-chip memory known as DMA and DRAM controller [6].”
MethodEscher: A CNN Accelerator with Flexible Buffering to Minimize Off-Chip Transfer
Shen, Y., Ferdman, M., & Milder, P. (2017) · 2017 IEEE 25th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM) · doi:10.1109/fccm.2017.47 · Cited by 106
García, G. J., Jara, C. A. V., Pomares, J., et al. (2014) · Sensors · doi:10.3390/s140406247 · Cited by 98
SqueezedText: A Real-Time Scene Text Recognition by Binary Convolutional Encoder-Decoder Network
Liu, Z., Li, Y., Ren, F., Goh, W. L., & Yu, H. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.12252 · Cited by 97
Hu, Q., Tang, X., & Tang, W. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.2980207 · Cited by 94 · cited 3×
“A FPGA with machine learning algorithm is then used to process the digital signal [65].”
MethodInfluentialALAMO: FPGA acceleration of deep learning algorithms with a modularized RTL compiler
Ma, Y., Suda, N., Cao, Y., Vrudhula, S., & Seo, J.-S. (2018) · Integration · doi:10.1016/j.vlsi.2017.12.009 · Cited by 93
A Survey and Taxonomy of FPGA-based Deep Learning Accelerators
Blaiech, A. G., Ben Khalifa, K., Valderrama, C., Fernandes, M. A. C., & Bedoui, M. H. (2019) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2019.01.007 · Cited by 88 · cited 1×
“Therefore, the comparison between CPU, GPU and FPGA had come into the picture in term of hardware resources, power consumption, design difficulty and accuracy [87].”
BackgroundDesigning efficient accelerator of depthwise separable convolutional neural network on FPGA
Ding, W., Huang, Z., Huang, Z., Tian, L., Wang, H., & Feng, S. (2019) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2018.12.008 · Cited by 85 · cited 7×
“Pipelined optimization technique [24] (Reprinted with permission from ref.”
BackgroundMethodInfluentialA highly scalable Restricted Boltzmann Machine FPGA implementation
Kim, S. K., McAfee, L. C., McMahon, P. L., & Olukotun, K. (2009) · 2009 International Conference on Field Programmable Logic and Applications · doi:10.1109/fpl.2009.5272262 · Cited by 80 · cited 4×
“The core block for the processing the GEMM is the processing element (PE) where all the multiplication and accumulation happen.”
MethodInfluentialMoss, D. J. M., Krishnan, S., Nurvitadhi, E., Ratuszniak, P., Johnson, C., Sim, J., Mishra, A., Marr, D., Subhaschandra, S., & Leong, P. H. W. (2018) · Proceedings of the 2018 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/3174243.3174258 · Cited by 78 · cited 1×
“The batch level parallelism is handled by PE which, in parallel, is using GEMM kernel [33].”
MethodTrimberger, S. M. S. (2018) · IEEE Solid-State Circuits Magazine · doi:10.1109/mssc.2018.2822862 · Cited by 77
FPGA-Accelerated Dense Linear Machine Learning: A Precision-Convergence Trade-Off
Kara, K., Alistarh, D., Alonso, G., Mutlu, O., & Zhang, C. (2017) · 2017 IEEE 25th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM) · doi:10.1109/fccm.2017.39 · Cited by 77 · cited 1×
“The authors in [8] gave a comparison of the SGD with quantization techniques in FPGA (in their approach using single-precision) having similar performance compared to 10-core CPU.”
ResultLi, S., Luo, Y., Sun, K., Yadav, N., & Choi, K. K. (2020) · IEEE Access · doi:10.1109/access.2020.3000009 · Cited by 75 · cited 1×
“In conclusion, large, reconfigurable processing elements with low power consumption have make FPGA widely used recently due to the rapid growth in demand in AI, so that 92% of the execution time is spent on MAC computations during forward inference in YOLO framework [85].”
MethodImplementing a fuzzy system on a field programmable gate array
McKenna, M. C., & Wilamowski, B. M. (2002) · doi:10.1109/ijcnn.2001.939015 · Cited by 73 · cited 1×
“A fuzzy system mainly consists of four main blocks such as a fuzzifier, rule base, inference engine and defuzzification [40].”
AddNet: Deep Neural Networks Using FPGA-Optimized Multipliers
Faraone, J., Kumm, M., Hardieck, M., Zipf, P., Liu, X., Boland, D., & Leong, P. H. W. (2020) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2019.2939429 · Cited by 66 · cited 2×
“The result of implementing RCCM and AddNet was the ability to save resources by up to 50% more than a standard 8-bit quantized network with similar accuracy [34].”
MethodMonotone Mamdani–Assilian models under mean of maxima defuzzification
Van Broekhoven, E., & De Baets, B. (2008) · Fuzzy Sets and Systems · doi:10.1016/j.fss.2008.03.014 · Cited by 63
An Overview of FPGA Based Deep Learning Accelerators: Challenges and Opportunities
Wang, T., Wang, C., Zhou, X., & Chen, H. (2019) · 2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th International Conference on Data Science and Systems (HPCC/SmartCity/DSS) · doi:10.1109/hpcc/smartcity/dss.2019.00229 · Cited by 53 · cited 2×
“In addition, challenges in programming are another increase in cost [88].”
BackgroundAn FPGA-based stream processor for embedded real-time vision with Convolutional Networks
Farabet, C., Poulet, C., & LeCun, Y. (2009) · 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops · doi:10.1109/iccvw.2009.5457611 · Cited by 50 · cited 1×
“The main architecture of OPF uses the single instruction multiple data (SIMD) approach with a group of processing elements (PE) and custom memory controllers for handling the data distribution given by the PE [14].”
MethodFrances-Villora, J. V., Rosado-Muñoz, A., Martínez-Villena, J. M., Bataller-Mompean, M., Guerrero, J. F., & Wegrzyn, M. (2016) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2016.02.007 · Cited by 49 · cited 2×
“Lastly, by applying the QR decomposition and pipelined architecture, it can save up to twenty-two times power consumption with similar performance as compared to the traditional ELM method [57].”
MethodLei, P., Liang, J., Guan, Z., Wang, J., & Zheng, T. (2019) · IEEE Access · doi:10.1109/access.2019.2926381 · Cited by 44 · cited 4×
“The human target-generated spectrogram of mmW radar mD signals is transferred and handled sequentially between layers [75].”
BackgroundMethodInfluentialFPGA-Based Convolutional Neural Network Architecture with Reduced Parameter Requirements
Hailesellasie, M., Hasan, S. R., Khalid, F., Wad, F. A., & Shafique, M. (2018) · 2018 IEEE International Symposium on Circuits and Systems (ISCAS) · doi:10.1109/iscas.2018.8351283 · Cited by 38 · cited 3×
“2 times fewer than standard network without impacting classification accuracy [28].”
BackgroundInfluentialA hardware accelerator for Particle Swarm Optimization
Calazan, R. M., Nedjah, N., & Mourelle, L. D. M. (2013) · Applied Soft Computing · doi:10.1016/j.asoc.2012.12.034 · Cited by 37 · cited 1×
“SI is optimized in FPGA by optimizing the baseline of the general framework of swarm intelligence (FASI) in FPGA by setting the tuning knob CN to 1 and R to 0 [51].”
MethodA Ubiquitous Machine Learning Accelerator with Automatic Parallelization on FPGA
Wang, C., Gong, L., Li, X., & Zhou, X. (2020) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2020.2990924 · Cited by 37 · cited 1×
“The authors in [5] proposed the Ubiquitous ML Accelerator with Automatic Parallelization on FPGA.”
MethodIntellino: Processor for Embedded Artificial Intelligence
Yoon, Y. H., Hwang, D. H., Yang, J. H., & Lee, S. E. (2020) · Electronics · doi:10.3390/electronics9071169 · Cited by 37 · cited 1×
“A recent work by the authors in [17] proposed a low-powered processor for embedded AI systems termed as Intellino.”
BackgroundA Fast and Scalable Architecture to Run Convolutional Neural Networks in Low Density FPGAs
Véstias, M. P., Duarte, R. P., de Sousa, J. T., & Neto, H. C. (2020) · Microprocessors and Microsystems · doi:10.1016/j.micpro.2020.103136 · Cited by 33 · cited 6×
“The 8-bit mixed fixed-point representation for weights and activation are able to improve performance and reduce the loss of accuracy in hardware [22].”
BackgroundMethodInfluentialChew, L. W., Chia, W. C., Ang, L.-M., et al. (2012) · International Journal of Sensor Networks · doi:10.1504/ijsnet.2012.045033 · Cited by 32
Self-citationA Wi-Fi-Based Wireless Indoor Position Sensing System with Multipath Interference Mitigation
Xie, T., Jiang, H., Zhao, X., & Zhang, C. (2019) · Sensors · doi:10.3390/s19183983 · Cited by 32
Ghaffari, A., & Savaria, Y. (2020) · Electronics · doi:10.3390/electronics9122200 · Cited by 31
Sousa, M. A. D. A. D., Pires, R., & Del-Moral-Hernandez, E. (2020) · Neural Networks · doi:10.1016/j.neunet.2020.02.019 · Cited by 30 · cited 4×
“The control unit is responsible for data transfer across the entire architecture [52].”
BackgroundMethodInfluentialParallel Implementation on FPGA of Support Vector Machines Using Stochastic Gradient Descent
Lopes, F. F., Ferreira, J. C., & Fernandes, M. A. C. (2019) · Electronics · doi:10.3390/electronics8060631 · Cited by 29
Kaarmukilan, S. P., Poddar, S., & Thomas K., A. (2020) · 2020 Fourth International Conference on Computing Methodologies and Communication (ICCMC) · doi:10.1109/iccmc48092.2020.iccmc-00088 · Cited by 29 · cited 2×
“Furthermore, there are other research works [61] to review about object detection in embedded intelligence.”
BackgroundImplementation and applications of tri-state self-organizing maps on FPGA
Appiah, K., Hunter, A., Dickinson, P., & Meng, H. (2012) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2012.2197077 · Cited by 28 · cited 3×
“distance to identify the best match units (BMU) [53], allowing for the use of high speed computation of the SOM learning phase by executing large SOM using a moderate density of FPGAs to speed up the data mining operations.”
MethodWireless Multimedia Sensor Networks on Reconfigurable Hardware: Information Reduction Techniques
Ang, L.-M., Seng, K. P., Chew, L. W., Yeong, L. S., & Chia, W. C. (2013) · Springer eBooks · doi:10.1007/978-3-642-38203-1 · Cited by 28
Self-citationLarge-Scale Deep Learning Framework on FPGA for Fingerprint-Based Indoor Localization
Liu, C., Wang, C., & Luo, J. (2020) · IEEE Access · doi:10.1109/access.2020.2985162 · Cited by 27
Towards efficient deep neural network training by FPGA-based batch-level parallelism
Luo, C., Sit, M.-K., Fan, H., Liu, S., Luk, W., & Guo, C. (2020) · Journal of Semiconductors · doi:10.1088/1674-4926/41/2/022403 · Cited by 26 · cited 1×
“Another proposed method, termed DarkFPGA by [32], uses batch-level parallelism hardware architecture.”
MethodA hybrid GPU-FPGA-based computing platform for machine learning
Liu, X., Ounifi, H. A., Gherbi, A., Lemieux, Y., & Li, W. (2018) · Procedia Computer Science · doi:10.1016/j.procs.2018.10.155 · Cited by 26 · cited 2×
“Based on the research, the FPGA Arria 10 speed is six times faster than GPU and 44 times faster than CPU in the inferencing process [83].”
MethodFPGA Platform applied for Facial Expression Recognition System using Convolutional Neural Networks
Phan-Xuan, H., Le-Tien, T., & Nguyen-Tan, S. (2019) · Procedia Computer Science · doi:10.1016/j.procs.2019.04.087 · Cited by 26 · cited 3×
“Another research done by [59] is using low cost computational FPGA Soc Zynq-7000 as platform for hardware accelerator which prove that it able significantly improve the real-time face recognition application.”
MethodInfluentialDesign guidelines for reconfigurable multiplier blocks
Demirsoy, S. S., Dempster, A. G., & Kale, I. (2003) · Proceedings of the 2003 International Symposium on Circuits and Systems, 2003. ISCAS '03. · doi:10.1109/iscas.2003.1205831 · Cited by 25 · cited 1×
“RCCM claims that it is a better alternative by replacing multiplication input values with only adders, subtractions, bit shifters and multiplexers [35]: for example, a constant coefficient multiplier in a circuit known as y = cx where c is constant.”
Deep Neural Network Model and FPGA Accelerator Co-Design: Opportunities and Challenges
Hao, C., & Chen, D. (2018) · 2018 14th IEEE International Conference on Solid-State and Integrated Circuit Technology (ICSICT) · doi:10.1109/icsict.2018.8564956 · Cited by 19 · cited 1×
“FPGA do not have built-in cache structure and the bandwidth typically is two generations older than high end GPU [89].”
BackgroundFPGA based nonlinear Support Vector Machine training using an ensemble learning
Rabieah, M. B., & Bouganis, C.-S. (2015) · International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl.2015.7293972 · Cited by 19 · cited 1×
“Support vector machines (SVM) are one of the most commonly used methods for machine learning that build a hyperplane to classify data between two classes [9], although a decomposition technique is commonly used to improve the performance of the SVM [10].”
MethodFrom CPU to FPGA — Acceleration of self-organizing maps for data mining
Lachmair, J., Mieth, T., Griessl, R., Hagemeyer, J., & Porrmann, M. (2017) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2017.7966400 · Cited by 19 · cited 1×
“Some researchers [54] have used multiple FPGA chips composed into single chipset to accelerate the SOM training process.”
MethodCoNNa–Hardware accelerator for compressed convolutional neural networks
Struharik, R. J. R., Vukobratović, B. Z., Erdeljan, A. M., & Rakanović, D. M. (2020) · Microprocessors and Microsystems · doi:10.1016/j.micpro.2020.102991 · Cited by 18 · cited 4×
“The authors in [18] proposed a hardware accelerator for compressed CNN (referred to as CoNNA) which is able to accelerate pruned, quantized CNN.”
BackgroundMethodInfluentialde Abreu de Sousa, M. A., & Del-Moral-Hernandez, E. (2017) · 2017 IEEE International Symposium on Circuits and Systems (ISCAS) · doi:10.1109/iscas.2017.8050799 · Cited by 18 · cited 1×
“In the paper by [55], focusing on centralized architecture due to low density, a FPGA chip is required to achieve higher processing speed.”
BackgroundConvolutional neural network for facial expression recognition
Lu, G., He, J., Yan, J., et al. (2016) · Journal of Nanjing University of Posts and Telecommunications · doi:10.14132/j.cnki.1673-5439.2016.01.003 · Cited by 18 · cited 1×
“The FER dataset with 48 × 48 pixel grayscale images of faces is used to test the FPGA implementation of the CNN method [60].”
MethodSwarm Intelligence Techniques and Their Adaptive Nature with Applications
Biswas, A., & Biswas, B. (2014) · Studies in fuzziness and soft computing · doi:10.1007/978-3-319-12883-2_9 · Cited by 15 · cited 2×
“At a later stage, a collection of algorithms is proposed to optimize problems such as particle swarm optimization (PSO) and ant colony optimization (ACO) [47].”
BackgroundAn Accelerated Prototype with Movidius Neural Compute Stick for Real-Time Object Detection
Kaarmukilan, S. P., Hazarika, A., Amal Thomas, K., Poddar, S., & Rahaman, H. (2020) · International Symposium on Devices, Circuits and Systems (ISDCS) · doi:10.1109/isdcs49393.2020.9262996 · Cited by 14 · cited 3×
“In term of frames per second (FPS) of the object detection model, SSD is the highest, with 3.98 FPS, while YOLO is the fastest for the computational time, with 12.68 ms. FRCNN is the most accurate, with 100% for probability of detection [61].”
MethodInfluentialCMOL implementation of spiking neurons and spike-timing dependent plasticity
Afifi, A., Ayatollahi, A., & Raissi, F. (2011) · International Journal of Circuit Theory and Applications · doi:10.1002/cta.638 · Cited by 13 · cited 1×
“ELM is typically used in single layer feedforward neural (SLFN) networks [56].”
MethodA fast FPGA-based deep convolutional neural network using pseudo parallel memories
Hailesellasie, M., & Hasan, S. R. (2017) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2017.8050317 · Cited by 13 · cited 1×
“All the layers that are removed will be replaced by a layer called the fire module layer [29] which is a combination of the squeeze layer and expand convolution layers.”
BackgroundOropharyngeal dysphagia identification using wavelets and optimum path forest
Spadotto, A. A., Pereira, J. C., Guido, R. C., Papa, J. P., Falcao, A. X., Gatto, A. R., Cola, P. C., & Schelp, A. O. (2008) · 2008 3rd International Symposium on Communications, Control and Signal Processing · doi:10.1109/isccsp.2008.4537320 · Cited by 12 · cited 1×
“Optimum path forest (OPF) is considered one of the graph-based classifiers quite similar to support vector machine (SVM) techniques but has a faster execution time and is easier to deploy [13].”
BackgroundResearch on high-speed fuzzy reasoning with FPGA for fault diagnosis expert system
Xue, J., Sun, L., Liu, M., Qiao, C., & Ye, G. (2009) · 2009 International Conference on Mechatronics and Automation · doi:10.1109/icma.2009.5245988 · Cited by 11 · cited 2×
“The high level of the expert system and the realizing process of the expert system with FPGA are discussed in [37].”
BackgroundDiniz, W. F. S., Fremont, V., Fantoni, I., & Nóbrega, E. G. O. (2017) · Microprocessors and Microsystems · doi:10.1016/j.micpro.2017.06.013 · Cited by 11 · cited 8×
“Figure 2 shows that each EPA consists of their own local shared memory (LSM) linked with a memory management (MM) module to protect the memory from corruption.”
BackgroundMethodInfluentialLi, D., Huang, L., Wang, K., Pang, W., Zhou, Y., & Zhang, R. (2018) · IEEE Access · doi:10.1109/access.2018.2882455 · Cited by 11 · cited 7×
“Figure 12 shows the rescheduled dataflow of quantum-behaved particle swarm optimization (QPSO).”
BackgroundMethodInfluentialJakšić, Z., Cadenelli, N., Prats, D. B., Polo, J., Berral Garcia, J. L., & Perez, D. C. (2020) · Future Generation Computer Systems · doi:10.1016/j.future.2019.10.025 · Cited by 10
ADAS and Video Surveillance Analytics System Using Deep Learning Algorithms on FPGA
Shan, Y. (2018) · 2018 28th International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl.2018.00092 · Cited by 10 · cited 2×
“Next, the facial recognition in the conference hall is performed by implementing the cameras to capture the real-time video, and then FPGA is used to analyze and display the result on the screen [73].”
MethodFPGA Implementation of Modified Gram-Schmidt QR-Decomposition
Ganchosv, P., Kusmanov., G., Kabakchiev, C. A., Behar, V. P., Romansky, R. P., & Gaydadjiev, G. N. (2009) · High Performance Embedded Architectures and Compilers · Cited by 9
A 307-fps 351.7-GOPs/W Deep Learning FPGA Accelerator for Real-Time Scene Text Recognition
Zhao, S., An, F., & Yu, H. (2019) · 2019 International Conference on Field-Programmable Technology (ICFPT) · doi:10.1109/icfpt47387.2019.00043 · Cited by 8 · cited 4×
“Encoder–decoder neural network [66] (Reprinted with permission from ref.”
BackgroundMethodInfluentialFPGA-based Accelerator for Independently Recurrent Neural Network
Gao, C., & Zhang, F. (2018) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/compcomm.2018.8780644 · Cited by 8
Reconfigurable Implementation of Fuzzy Inference System Using FPGA
Mohammadi, M., & Shaout, A. (2017) · doi:10.1109/ictcs.2017.63 · Cited by 5
FPGA Implementation of a Fuzzy Rule Based Contrast Enhancement System for Real Time Applications
Liviu, T. (2018) · 2018 22nd International Conference on System Theory, Control and Computing (ICSTCC) · doi:10.1109/icstcc.2018.8540721 · Cited by 4 · cited 5×
“Fuzzy core block main structure [44] (Reprinted with permission from ref.”
BackgroundMethodInfluentialFuzzy Vs. neuro-fuzzy: Implementation on the reconfigurable FPGA system
Mondal, S., & Chattapadhyay, P. (2015) · doi:10.1109/epetsg.2015.7510074 · Cited by 2 · cited 3×
“• COG strength is the capability to accelerate the CMOS arithmetic architecture.”
MethodDesign of a blowout expert control system based on FPGA
Da-cheng, M., & Yu-h, L. (2010) · 2010 The 2nd International Conference on Computer and Automation Engineering (ICCAE) · doi:10.1109/iccae.2010.5451650 · Cited by 1 · cited 2×
“The reasoning mechanism based on the production rules is shown as U = f(E, K, I, G) where E is the input information, K is the empirical fact and U is the output of control behaviour [38].”
BackgroundMethod
Research works printed but not linked: 3
- Duan, 2006, Fuzzy-PID control of stepping motor, Comput. Simul., 2, 290
- Eberchart, J.K.R. (December, January 27). Particle Swarm Optimization. Proceedings of the IEEE International Conference on Neural Networks, Perth, Australia.
- Xilinx (2019, April 24). Xilinx Reports Record Revenues Exceeding $3 Billion For Fiscal. Available online: https://www.xilinx.com/news/press/2019/xilinx-reports-record-revenues-exceeding-3-billion-for-fiscal-2019.html.
Cited by
Lightweight Deep Learning for Resource-Constrained Environments: A Survey
Liu, H.-I., Galindo, M., Xie, H., Wong, L.-K., Shuai, H.-H., Li, Y.-H., & Cheng, W.-H. (2024) · ACM Computing Surveys · doi:10.1145/3657282 · Cited by 259
“For more references on FPGA AI applications, consult References [148, 170].”
BackgroundUsing acoustic indices in ecology: Guidance on study design, analyses and interpretation
Bradfer‐Lawrence, T., Desjonqueres, C., Eldridge, A., Johnston, A., & Metcalf, O. (2023) · Methods in Ecology and Evolution · doi:10.1111/2041-210x.14194 · Cited by 106
Machine Learning for the Control and Monitoring of Electric Machine Drives: Advances and Trends
Zhang, S., Wallscheid, O., & Porrmann, M. (2023) · IEEE Open Journal of Industry Applications · doi:10.1109/ojia.2023.3284717 · Cited by 122
“of the weights is reduced, resulting in reduced model size and faster computation at the cost of minimal reductions in predictive accuracy [79].”
BackgroundA review of AI edge devices and lightweight CNN and LLM deployment
Sun, K., Wang, X., Miao, X., & Zhao, Q. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128791 · Cited by 103
Disclosing Edge Intelligence: A Systematic Meta-Survey
Barbuto, V., Savaglio, C., Chen, M., & Fortino, G. (2023) · Big Data and Cognitive Computing · doi:10.3390/bdcc7010044 · Cited by 80
“However, they refer to specific domains (e.g., robotics [18] or cyber-security [19]) or technology (e.g., embedded intelligence for FPGA [20]), to narrow contributions (e.g., a novel version of an algorithm or an optimized model [21,22]), and to broader surveys/review papers outlining definitions,…”
Metadata
- Authors
- Seng, K. P., Lee, P. J., & Ang, L. M.
- Affiliations
- UNSW Sydney (AU)
- University of the Sunshine Coast (AU)
- Corresponding author
- Seng, Kah Phooi
- Venue
- Electronics
- Publisher
- MDPI AG
- Published
- 2021-04-08
- Volume
- 10
- Issue
- 8
- Pages
- 895
- ISSN
- 2079-9292 (ISSN-L)
- OpenAlex
- W3148609646
- Cited by
- 105
- References
- 93
- Citations by year
- 2026: 14 · 2025: 18 · 2024: 25 · 2023: 19 · 2022: 21 · 2021: 8
- FWCI
- 11.97
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Electronics (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Embedded Systems and FPGA Applications
- Physical Sciences › Engineering › Control and Systems Engineering › Embedded Systems and FPGA Design
- Keywords
- embedded intelligence, FPGA, low power consumption, energy efficiency, scalability, artificial neural networks, deep learning, expert systems, swarm intelligence, self-organizing maps, object detection and recognition, indoor positioning, surveillance monitoring, embedded devices
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-06, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-06 (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
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.