CNP: An FPGA-based processor for Convolutional Networks
The DOI matches its Crossref record; 16 of the 17 references it cites were found in a registry or an index.
Abstract
Convolutional networks (ConvNets) are biologically inspired hierarchical architectures that can be trained to perform a variety of detection, recognition and segmentation tasks. ConvNets have a feed-forward architecture consisting of multiple linear convolution filters interspersed with pointwise non-linear squashing functions. This paper presents an efficient implementation of ConvNets on a low-end DSP-oriented field programmable gate array (FPGA). The implementation exploits the inherent parallelism of ConvNets and takes full advantage of multiple hardware multiply accumulate units on the FPGA. The entire system uses a single FPGA with an external memory module, and no extra parts. A network compiler software was implemented, which takes a description of a trained ConvNet and compiles it into a sequence of instructions for the ConvNet Processor (CNP). A ConvNet face detection system was implemented and tested. Face detection on a 512 times 384 frame takes 100 ms (10 frames per second), which corresponds to an average performance of 3.4 times 109connections per second for this 340 million connection network. The design can be used for low-power, lightweight embedded vision systems for micro-UAVs and other small robots.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Another option is to use field programable gate arrays (FPGAs), which consume much less energy than GPUs while providing an intermediate computing efficiency between CPUs and GPUs13–17.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [13]
In the citing paper
“Also, CNN can be deployed in medical image classification [114], [115], traffic scenes related classification [116], [117], etc. [118], [119]”
From A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospects (2022) · cites this paper as [116]
Citations
Cites
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 4×
“The design is specifically geared towards Convolutional Networks [1, 2], but can be used for many similar architectures based on local filter banks and classifiers, such as HMAX [3, 4], and HoG methods [5].”
BackgroundMethodInfluentialSelf-citationHistograms of Oriented Gradients for Human Detection
Dalal, N., & Triggs, B. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.177 · Cited by 33,584 · cited 1×
“The design is specifically geared towards Convolutional Networks [1, 2], but can be used for many similar architectures based on local filter banks and classifiers, such as HMAX [3, 4], and HoG methods [5].”
MethodRapid object detection using a boosted cascade of simple features
Viola, P., & Jones, M. (2005) · Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001 · doi:10.1109/cvpr.2001.990517 · Cited by 18,300
GradientBased Learning Applied to Document Recognition
(2009) · Intelligent Signal Processing · doi:10.1109/9780470544976.ch9 · Cited by 1,135 · cited 5×
“This paper presents a self-contained, high performance implementation of Convolutional Networks on a single FPGA.”
BackgroundMethodInfluentialUnsupervised Learning of Invariant Feature Hierarchies with Applications to Object Recognition
Ranzato, M., Huang, F. J., Boureau, Y.-L., & LeCun, Y. (2007) · 2007 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2007.383157 · Cited by 1,086 · cited 1×
“ConvNets are used in several commercial and experimental applications, including video surveillance, OCR [2], face/person detection [6, 7], object recognition [8], and robot navigation [9, 10].”
MethodSelf-citationGeneralization and network design strategies
LeCun, Y. (1989) · Cited by 885
Self-citationObject Recognition with Features Inspired by Visual Cortex
Serre, T., Wolf, L., & Poggio, T. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.254 · Cited by 821 · cited 1×
“The design is specifically geared towards Convolutional Networks [1, 2], but can be used for many similar architectures based on local filter banks and classifiers, such as HMAX [3, 4], and HoG methods [5].”
MethodConvolutional face finder: a neural architecture for fast and robust face detection
Garcia, C., & Delakis, M. (2004) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2004.97 · Cited by 547 · cited 3×
“The architecture of the network is quite similar to those described in [6, 7].”
MethodMulticlass Object Recognition with Sparse, Localized Features
Mutch, J., & Lowe, D. G. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 1 (CVPR'06) · doi:10.1109/cvpr.2006.200 · Cited by 453 · cited 1×
“The design is specifically geared towards Convolutional Networks [1, 2], but can be used for many similar architectures based on local filter banks and classifiers, such as HMAX [3, 4], and HoG methods [5].”
Method(2005) · neural information processing systems · Cited by 446
Self-citation
Research works printed but not linked: 1
- shoup, 1994, parameterized convolution filtering in a field programmable gate array, Selected papers from the Oxford 1993 international workshop on field programmable logic and applications on More FPGAs, 274
Cited by
Items 201–300 of 423
Towards real-time image understanding with convolutional networks
Farabet, C. (2013) · theses.fr (ABES) · doi:10.70675/98712c00z2d54z4479z8e3bz0716b256db97 · Cited by 13
Self-citationHoshi, I., Wakunami, K., Ichihashi, Y., & Oi, R. (2025) · Scientific Reports · doi:10.1038/s41598-024-82791-z · Cited by 12
Exploring Portability and Performance of OpenCL FPGA Kernels on Intel HARPv2
Cabrera, A. M., & Chamberlain, R. D. (2019) · International Workshop on OpenCL · doi:10.1145/3318170.3318180 · Cited by 12
“Additionally, there is a growing research trend toward harnessing the reconfigurability of FPGAs towards salient applications like accelerating neural networks [5, 11, 30], biocompu-tation [15, 18, 19], and many other applications [17, 24, 26, 29, 34].”
BackgroundDeep Learning in Bioinformatics
(2022) · Elsevier eBooks · doi:10.1016/c2020-0-00432-9 · Cited by 12
AivoTTA: an energy efficient programmable accelerator for CNN-based object recognition
IJzerman, J., Viitanen, T. T., Jääskeläinen, P., Kultala, H. O., Lehtonen, L. K., Peemen, M., Corporaal, H., & Takala, J. (2018) · International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation · doi:10.1145/3229631.3229637 · Cited by 11
“Implementations on FPGAs [14, 37, 16, 29] offer great flexibility with limited performance.”
BackgroundFeatherNet: An Accelerated Convolutional Neural Network Design for Resource-constrained FPGAs
Morcel, R., Hajj, H. M., Saghir, M. A. R., Akkary, H. H., Artail, H. A., Khanna, R., & Keshavamurthy, A. S. (2019) · ACM Transactions on Reconfigurable Technology and Systems · doi:10.1145/3306202 · Cited by 11
“Moreover, the ConvNets that were deployed in References [10–12] do not contain many of the more intricate structures and layers involved in current state-of-art ConvNets.”
BackgroundMethodInfluentialLou, Q., Wen, W., & Jiang, L. (2018) · doi:10.1145/3240765.3240767 · Cited by 11
“We studied six CNNs including LeNet-5 [20], CNP [11], SCNN [29], MCDNN [7], AlexNet [19] and ResNet-18 [15], and a Multilayer Perceptron(MLP) [20].”
MethodDeep-Learning Inferencing with High-Performance Hardware Accelerators
Kljucaric, L., & George, A. (2019) · IEEE Conference on High Performance Extreme Computing · doi:10.1109/hpec.2019.8916463 · Cited by 46
“For more information on previous work accelerating CNNs on FPGAs, the reader is referred to [2], [3], [4], [5], [30].”
MethodData Optimization CNN Accelerator Design on FPGA
Hu, W., Chen, S., Li, Z., Liu, T., & Li, Y. (2019) · IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA) · doi:10.1109/ispa-bdcloud-sustaincom-socialcom48970.2019.00051 · Cited by 11
“In the last few years, convolutional neural network (CNN) have been adopted in various applications, which include mobile robot vision, video surveillance, image search engine in data centers, etc. [5, 6, 7, 8, 9].”
BackgroundA Novel Approach to Develop Low Power MACs for 2D Image Filtering
Samanth, R., Vishnumurthy Kedlaya, K., & Nayak, S. G. (2021) · IEEE Access · doi:10.1109/access.2021.3058736 · Cited by 10
“The conventional design was limited to window size 3 × 3 for 2D convolution [4].”
MethodInfluentialONNX-To-Hardware Design Flow for Adaptive Neural-Network Inference on FPGAs
Manca, F., Ratto, F., & Palumbo, F. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-78380-7_7 · Cited by 10
An Energy-Efficient Silicon Photonic-Assisted Deep Learning Accelerator for Big Data
Li, M., & Wang, Y. (2020) · Wireless Communications and Mobile Computing · doi:10.1155/2020/6661022 · Cited by 10
Efficient Zero-Activation-Skipping for On-Chip Low-Energy CNN Acceleration
Liu, M., He, Y., & Jiao, H. (2021) · IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS) · doi:10.1109/aicas51828.2021.9458578 · Cited by 10
“Even though having no effect on the final result, these zeros still occupy MACs in the traditional convolution calculation method [8], which makes limited hardware resources less available.”
BackgroundMethodVarghese, A., Sylaja, M. M., & Kurian, J. (2022) · Journal of Intelligent Systems · doi:10.1515/jisys-2022-0015 · Cited by 10
CASA: A Convolution Accelerator using Skip Algorithm for Deep Neural Network
Kim, Y. H., An, G. J., & Sunwoo, M. H. (2019) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2019.8702307 · Cited by 10
A Platform based on HLS to Implement a Generic CNN on an FPGA
Baptista, D., Morgado‐Dias, F., & Sousa, L. A. (2019) · International Conference in Engineering Applications (ICEA) · doi:10.1109/ceap.2019.8883473 · Cited by 10
“In the literature other similar approaches are presented [6]–[10].”
BackgroundMethodHigh Throughput Spatial Convolution Filters on FPGAs
Ioannou, L., Al-Dujaili, A., & Fahmy, S. A. (2020) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2020.2987202 · Cited by 10
“[32] 7×7 V4SX35 200 - - Neural Network related”
You, Y., Chang, Y., Wu, W., Guo, B., Luo, H., Liu, X., Liu, B., Zhao, K., He, S., Li, L., et al. (2022) · Design Automation for Embedded Systems · doi:10.1007/s10617-021-09256-8 · Cited by 10
LTE downlink throughput modeling using neural networks
Rehman, T. U., Baig, M. A. I., & Ahmad, A. (2017) · IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON) · doi:10.1109/uemcon.2017.8249044 · Cited by 10
“Since 2010, neural networks have found overwhelming applications in tasks that involve prediction, forecasting and pattern recognition, which including stock price estimation [8], electricity load forecasting [9], space exploration [10], driving autonomous vehicles [11],[12], natural language processing [13],image rec…”
BackgroundScaling Binarized Neural Networks on Reconfigurable LogicPreprint
Fraser, N. J., Umuroglu, Y., Gambardella, G., Blott, M., Leong, P. H. W., Jahre, M., & Vissers, K. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1701.03400 · Cited by 10
A Tale of Two Domains: Exploring Efficient Architecture Design for Truly Autonomous Things
Hou, X., Xu, T., Li, C., Xu, C., Liu, J., Hu, Y., Zhao, J., Leng, J., Cheng, K.-T. T., & Guo, M. (2024) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca59077.2024.00022 · Cited by 9
“There have been several efforts to explore acceleration of ML inference with MB/GB sized models on FPGAs [7], [18], [21].”
Rupanetti, D., Nepal, K., SALAMY, H. A., & Min, C.-H. (2020) · IEEE Annual Computing and Communication Workshop and Conference (CCWC) · doi:10.1109/ccwc47524.2020.9031161 · Cited by 9
An implementation methodology for Neural Network on a Low-end FPGA Board
Wei, K., Honda, K., & Amano, H. (2020) · International Symposium on Computing and Networking · doi:10.1109/candar51075.2020.00039 · Cited by 9
“The benefit of using FPGAs for We compared our work with some other low-end FPGA implementation, which includes DLAU [19] with the precision of 48-bit floating-point, 16-bit fixed-point CNP [20], and FP-GAConvNet [21].”
MethodBest Practices for the Deployment of Edge Inference: The Conclusions to Start Designing
Flamis, G., Kalapothas, S., & Kitsos, P. (2021) · Electronics · doi:10.3390/electronics10161912 · Cited by 9
Image Quantization Tradeoffs in a YOLO-based FPGA Accelerator Framework
Yarnell, R., Hossain, M. S. B., & DeMara, R. F. (2023) · International Symposium on Quality Electronic Design (ISQED) · doi:10.1109/isqed57927.2023.10129324 · Cited by 9
“Background and Related Work on FPGA-based CNN Accelerators The first implementations of an FPGA-based neural network dates to 2009, where Farabet et al implemented a ConvNet-based CNN processor on an FPGA for a facial recognition task on a low-cost robot [3].”
MethodAn Accelerator for High Efficient Vision Processing
Du, Z., Liu, S., Fasthuber, R., Chen, T., Ienne, P., Li, L., Luo, T., Guo, Q., Feng, X., Chen, Y., et al. (2016) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2016.2584062 · Cited by 9
“We collected 10 CNNs from representative visual recognition applications and used them as our benchmarks (CNP [46], MPCNN [43], Face Recog.”
BackgroundMethodHardware Design Automation of Convolutional Neural Networks
Solazzo, A., Sozzo, E. D., Rose, I. D., Silvestri, M. D., Durelli, G. C., & Santambrogio, M. D. (2016) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi.2016.101 · Cited by 9
“Farabet et al. presented a programmable CNN processor implemented on a low-end DSP-oriented FPGA [13].”
BackgroundMethodOn How to Design Dataflow FPGA-Based Accelerators for Convolutional Neural Networks
Natale, G., Bacis, M., & Santambrogio, M. D. (2017) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi.2017.126 · Cited by 9
“Research has therefore focused on exploring specialized hardware accelerators based on GPUs, ASICs and FPGAs, achieving superior performance figures, with a better energyefficiency[5, 20, 16, 15].”
MethodNetwork Enabled Partial Reconfiguration for Distributed FPGA Edge Acceleration
Bucknall, A. R., Shreejith, S., & Fahmy, S. A. (2019) · International Conference on Field-Programmable Technology (FPT) · doi:10.1109/icfpt47387.2019.00042 · Cited by 9
“The use of FPGAs within data centres, has enabled efficient acceleration of operations such as data/feature extraction, dataflow management and information analysis [1], for a range of applications including image recognition [2], in-network security [3] and machine learning [4], among others.”
MethodHigh Utilization Energy-Aware Real-Time Inference Deep Convolutional Neural Network Accelerator
Lin, K.-T., Chiu, C., Chang, J.-Y., & Hsiao, S.-C. (2021) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas51556.2021.9401526 · Cited by 9
Ultra Low Latency Machine Learning for Scientific Edge Applications
Miniskar, N. R., Young, A. R., Liu, F., Blokland, W., Cabrera, A. M., & Vetter, J. S. (2022) · International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl57034.2022.00068 · Cited by 9
“An incomplete list of representative papers includes [14], [15], [16], [17], [18].”
BackgroundTraining Bit Fully Convolutional Network for Fast Semantic SegmentationPreprint
Wen, H., Zhou, S., Liang, Z., Zhang, Y., Feng, D., Zhou, X., & Yao, C. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1612.00212 · Cited by 9
An Efficient Deep Network in Network Architecture for Image Classification on FPGA Accelerator
Alaeddine, H., Malek, J., & Khemaja, M. (2021) · International Conference on Cyberworlds (CW) · doi:10.1109/cw52790.2021.00018 · Cited by 8
A Locality Aware Convolutional Neural Networks Accelerator
Shi, R., Xu, Z. G., Sun, Z., Peemen, M., Li, A., Corporaal, H., & Wu, D. (2015) · Euromicro Conference on Digital System Design (DSD) · doi:10.1109/dsd.2015.70 · Cited by 8
“To meet the high performance requirements, typical CNN solutions fall into three categories: 1) mapping to GPUs [6] [5], 2) acceleration by FPGAs [11] [17] [14] [15] and 3) accelerators with ASIC technology [10] [12] [13].”
MethodAn FPGA-based accelerator for cortical object classification
Park, M. S., Kestur, S., Sabarad, J., Narayanan, V., & Irwin, M. J. (2012) · Design, Automation, and Test in Europe · doi:10.5555/2492708.2492884 · Cited by 8
Flexible Fixed-Point Function Generation for FPGAs
Istoan, M., & Pasca, B. (2017) · IEEE Symposium on Computer Arithmetic (ARITH) · doi:10.1109/arith.2017.31 · Cited by 8
“TABLE II PERFORMANCE FOR SOME OF DESCRIBED ARCHITECTURES ON CYCLONEV C6 FOR X ∈ [1,2).”
BackgroundMethodInfluentialAn efficient FPGA-Based architecture for convolutional neural networks
Hwang, W., Jhang, Y.-J., & Tai, T. (2017) · International Conference on Telecommunications and Signal Processing (TSP) · doi:10.1109/tsp.2017.8076054 · Cited by 8
“A number of FPGA-based implementations have been proposed [4], [5], [6], [7], [8], [9], [10] for the classification phase of a CNN.”
BackgroundMethodInfluentialAccelerating Deep Neural Network Computation on a Low Power Reconfigurable Architecture
Xiong, Y., Zhou, J., Pal, S., Blaauw, D. T., Kim, H. S., Mudge, T., Dreslinski, R., & Chakrabarti, C. (2020) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas45731.2020.9180871 · Cited by 8
“An example is field programmable gate arrays (FPGA) which have been shown to provide high efficiency on machine learning [10], [11] and DNN [12] algorithms but at the expense of large area and power consumption.”
BackgroundRaising the Abstraction Level of a Deep Learning Design on FPGAs
Baptista, D., Sousa, L. A., & Morgado‐Dias, F. (2020) · IEEE Access · doi:10.1109/access.2020.3036975 · Cited by 8
Optimizing convolutional neural network on FPGA under heterogeneous computing framework with OpenCL
Wang, Z., Qiao, F., Liu, Z., Shan, Y., Zhou, X., Luo, L., & Yang, H. (2016) · IEEE Region 10 Conference (TENCON) · doi:10.1109/tencon.2016.7848692 · Cited by 8
“INTRODUCTION Convolutional neural network (CNN) has been widely used in many applications, including human face recognition, image classification, artificial intelligence, etc [1].”
BackgroundA Hardware Accelerator for Convolutional Neural Network Using Fast Fourier Transform
Kala, S., Jose, B. R., Paul, D., & Mathew, J. (2019) · Communications in computer and information science · doi:10.1007/978-981-13-5950-7_3 · Cited by 8
Hardware Acceleration of Convolutional Neural Networks
Halvorsen, M. (2015) · NORA - Norwegian Open Research Archives · Cited by 8
A survey of neural network accelerator with software development environments
Song, J. Y., Wang, X., Zhao, Z., Li, W., & Zhi, T. (2020) · Journal of Semiconductors · doi:10.1088/1674-4926/41/2/021403 · Cited by 7
Bearing Fault Identification Based on Deep Convolution Residual Network
Zhou, T., Li, Y., Yijia, J. I. N. G., & Tong, Y. (2021) · Mechanika · doi:10.5755/j02.mech.28265 · Cited by 7
NnCore: A parameterized non-linear function generator for machine learning applications in FPGAs
Ho, S. M. H., & So, H. K. (2017) · International Conference on Field-Programmable Technology (FPT) · doi:10.1109/fpt.2017.8280134 · Cited by 7
“piecewise linear approximate with number of segments from empirical experience [4][5][6][7], without quantifying the consequence on the networks’ accuracy.”
BackgroundENAS oriented layer adaptive data scheduling strategy for resource limited hardware
Li, C., Fan, X., Geng, Y., Yang, Z., Wang, D., & Zhang, M. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.11.005 · Cited by 7
An Embedded Inference Framework for Convolutional Neural Network Applications
Bi, S., Zhang, Y., Dong, M., & Min, H. (2019) · IEEE Access · doi:10.1109/access.2019.2956080 · Cited by 7
“In [18], an efficient CNN and an efficient CNN-based face detection system are implemented using a single FPGA by Farabet et al.”
MethodMinimal Filtering Algorithms for Convolutional Neural Networks
Cariow, A., & Cariowa, G. (2021) · Studies in computational intelligence · doi:10.1007/978-3-030-74556-1_5 · Cited by 7
Wang, Z., Li, H., Yue, X., & Meng, L. (2022) · Applied Sciences · doi:10.3390/app12147337 · Cited by 7
“Implementing deep learning using a field programmable gate array (FPGA) [34] can accelerate deep learning, reduce energy consumption, and deploy CNN on edge platforms and embedded systems.”
MethodPerformance-oriented FPGA-based convolution neural network designs
Kao, C. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-14537-4 · Cited by 7
“After completing the training, the characteristics of the set of data had been learned, and the content of the undetermined image can be known by using these features [8, 14, 22].”
BackgroundOptimizing CNNs on Multicores for Scalability, Performance and Goodput
Rajbhandari, S., He, Y., Ruwase, O., Carbin, M., & Chilimbi, T. M. (2017) · ACM SIGARCH Computer Architecture News · doi:10.1145/3093337.3037745 · Cited by 7
Coarse-grained Dynamically Reconfigurable Processor for Vision Pre-Processing
Zhang, B., Mei, K., & Zheng, N. (2013) · Journal of Signal Processing Systems · doi:10.1007/s11265-013-0828-1 · Cited by 7
“A 250MHz convolution network processor implemented on a DSP-oriented FPGA has achieved the peak processing performance of 24.5GOPS [12].”
MethodFPGA Implementation of Face Recognition System Based on Convolution Neural Network
Qiao, S., & Ma, J. (2018) · Chinese Automation Congress · doi:10.1109/cac.2018.8623662 · Cited by 7
“An architecture for CNN is proposed in [9] and the CNN is tested by running a face detection system on it.”
MethodConfigurable N-fold Hardware Architecture for Convolutional Neural Networks
Baptista, D., Morgado‐Dias, F., & Sousa, L. A. (2018) · International Conference on Biomedical Engineering and Applications (ICBEA) · doi:10.1109/icbea.2018.8471739 · Cited by 6
“[16] proposed a different implementation of a CNN on an FPGA, which can achieve 10 GOPs (Giga Operations per second).”
MethodLi, J., Yan, G., Lü, W., Gong, S., Jiang, S., Wu, J., & Li, X. (2018) · ACM Transactions on Design Automation of Electronic Systems · doi:10.1145/3275243 · Cited by 6
“Neural networks have been widely used in various fields such as image classification (Dahl et al. 2013; Farabet et al. 2009; Huang et al. 2013; Ji et al. 2013; Larochelle et al. 2007; Mnih and Hinton 2012).”
BackgroundMethodA Z Structure Convolutional Neural Network Implemented by FPGA in Deep Learning
Zhu, M., Kuang, Q., Lin, J., Luo, Q., Yang, C., & Liu, M. (2018) · Annual Conference of the IEEE Industrial Electronics Society (IECON) · doi:10.1109/iecon.2018.8592775 · Cited by 6
“In recent years, with the rapid development of high performance FPGA, the researchers pay more attention on low power consumption and flexible architecture design([10,11,12]).”
BackgroundE3: A HW/SW Co-design Neuroevolution Platform for Autonomous Learning in Edge Device
Kao, S.-C., & Krishna, T. (2021) · IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS) · doi:10.1109/ispass51385.2021.00051 · Cited by 6
“Also, many FPGA-based accelerator, featuring its flexibility, were proposed, such as Caffeine [48], Cnp [10], and others [23], [11], [47], [39], [50].”
MethodAhmad, H., Tanvir, M., Hanif, M. A., Javed, M. U., Hafiz, R., & Shafique, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.04986 · Cited by 6
“Recent advances in their hardware acceleration has mostly focused on maximizing the design reuse [3] for weights [4][5] and/or the input/output feature maps [6][7].”
BackgroundReconfigurable hardware acceleration of CNNs on FPGA-based smart cameras
Abdelouahab, K. (2018) · theses.fr (ABES) · doi:10.70675/88c8fb7dz3228z4c3cz926cz62f365d6d2d7 · Cited by 6
Area and Power Efficient Design of Novel Karatsuba Double MAC (K-DMAC)
Laxman, A., Reddy, N. S. S., & Naik, B. R. (2022) · International Conference on Inventive Research in Computing Applications (ICIRCA) · doi:10.1109/icirca54612.2022.9985758 · Cited by 5
Unified Virtual Memory Support for Deep CNN Accelerator on SoC FPGA
Xiao, T., Qiao, Y., Shen, J., Yang, Q., & Wen, M. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-27119-4_5 · Cited by 5
Limonova, E. E., Bocharov, N., Paramonov, N., Bogdanov, D. S., Arlazarov, V. V., Slavin, O., & Nikolaev, D. (2019) · Programming and Computer Software · doi:10.1134/s0361768819010055 · Cited by 5
“The performance of these systems can be improved by using programmable logic devices (PLDs) to implement the entire system or its individual modules [5, 6].”
BackgroundOn-the-Fly Lowering Engine: Offloading Data Layout Conversion for Convolutional Neural Networks
Kang, M., Hyun, S., Han, T. H., Kim, J., & Hong, S. (2022) · IEEE Access · doi:10.1109/access.2022.3192618 · Cited by 5
Design of a Lightweight Convolutional Neural Network Accelerated by FPGA
Ying-jun, X., & Wang, K. (2024) · International Conference on Algorithm, Image Processing and Machine Vision (AIPMV) · doi:10.1109/aipmv62663.2024.10692091 · Cited by 5
“To deploy a CNN model on the PL side of the FPGA, it is typically necessary to convert floating-point numbers to fixed-point numbers, as the fixed-point parameters can effectively save hardware resources [8].”
MethodDLPlib: A Library for Deep Learning Processor
Lan, H., Wu, L., Zhang, X., Tao, J., Chen, X.-Y., Wang, B.-R., Wang, Y., Guo, Q., & Chen, Y. (2017) · Journal of Computer Science and Technology · doi:10.1007/s11390-017-1722-2 · Cited by 5
Low-Power U-Net for Semantic Image Segmentation
Narendra, V., Peesapati, R., & Balabantaray, B. K. (2021) · Lecture notes in electrical engineering · doi:10.1007/978-981-16-0289-4_35 · Cited by 5
Real-Time Information Fusion System Implementation Based on ARM-Based FPGA
Tsai, Y.-H., Yan, Y., Hsiao, M.-H., Yu, T., & Ou‐Yang, M. (2023) · Applied Sciences · doi:10.3390/app13148497 · Cited by 5
Colodro-Conde, C., Toledo, F. J., Toledo-Moreo, R., Martínez-Álvarez, J. J., Garrigós, J., & Vicente, J. M. F. (2013) · Journal of Real-Time Image Processing · doi:10.1007/s11554-013-0374-x · Cited by 5
“2-D convolution is also the basis of machine learning algorithms [27] or biologically-inspired models [12, 22, 29].”
BackgroundLü, W., Yan, G., Li, J., Gong, S., Jiang, S., Wu, J., & Li, X. (2018) · IEEE Transactions on Computers · doi:10.1109/tc.2018.2890258 · Cited by 5
“To make it computationally affordable, the state-of-the-art optimizations mainly focus on two types of fundamental approaches: 1) accelerating computation by exploiting intensive parallelisms, such as TPU [11], DC-CNN [12], CNP [13], NeuFlow [14], SmartShuttle [15], NNest [16], Diannao family [17], [18], [19], [20], E…”
MethodJadhav, S. S., Gloster, C., Doss, C., Kim, Y.-S., & Naher, J. (2018) · IEEE International Performance, Computing, and Communications Conference · doi:10.1109/pccc.2018.8710782 · Cited by 5
“Applications mapped to FPGAs include: finance [4], machine learning [5], image processing [6, 7], graph analysis algorithms [1], pattern recognition [8], etc.”
Comparative Study of Hardware Accelerated Convolution Neural Network on PYNQ Board
Salman, A. M., Tulan, A. S., Mohamed, R. Y., Zakhari, M. H., & Mostafa, H. (2020) · Novel Intelligent and Leading Emerging Sciences Conference (NILES) · doi:10.1109/niles50944.2020.9257899 · Cited by 5
“RTL is a mature technology with a well-experienced developers and wide range of maintained IPs that you can use [11], [12].”
BackgroundOptimizing CNNs on Multicores for Scalability, Performance and Goodput
Rajbhandari, S., He, Y., Ruwase, O., Carbin, M., & Chilimbi, T. M. (2017) · ACM SIGOPS Operating Systems Review · doi:10.1145/3093315.3037745 · Cited by 5
Cariow, A., & Cariowa, G. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.03458 · Cited by 5
“In low power application specific integrated circuits (ASIC) design, optimization must be primarily done at the level of transistor amount.”
MethodInfluentialDesign and Implementation of Hardware Accelerators for Neural Processing ApplicationsPreprint
Mayannavar, S., & Wali, U. V. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.00051 · Cited by 5
Sudharsan, B., Sundaram, D., Patel, P., Breslin, J. G., Ali, M. I., Dustdar, S., Zomaya, A. Y., & Ranjan, R. (2022) · arXiv (Cornell University) · doi:10.13140/rg.2.2.35307.62244 · Cited by 5
Lightweight deep neural network based fault detection and diagnosis for quadrotor propeller faults
Adaika, Z., & Boumehraz, M. (2025) · International Journal of Intelligent Robotics and Applications · doi:10.1007/s41315-025-00500-z · Cited by 4
Enabling Efficient Training of Convolutional Neural Networks for Histopathology Images
Alali, M. H., Roohi, A., & Deogun, J. S. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-13321-3_47 · Cited by 4
Young, A. R., Miniskar, N. R., Liu, F., Blokland, W., & Vetter, J. S. (2022) · Communications in computer and information science · doi:10.1007/978-3-031-23606-8_14 · Cited by 4
Yu, Y., Virupakshappa, K., & Oruklu, E. (2023) · Journal of Signal Processing Systems · doi:10.1007/s11265-023-01901-8 · Cited by 4
“One of the earlier works of deep learning implementation on FPGA is [10].”
Frames-to-AER efficiency study based on CPUs performance counters
Domínguez-Morales, M., Iñigo-Blasco, P., Font, J. L., Cascado, D., Jiménez, G., Díaz, F., Sevillano, J. L., & Linares-Barranco, A. (2010) · idUS (Universidad de Sevilla) · Cited by 4
Jet Features: Hardware-Friendly, Learned Convolutional Kernels for High-Speed Image Classification
Simons, T., & Lee, D.-J. (2019) · Electronics · doi:10.3390/electronics8050588 · Cited by 4
Hardware Implementation of Reconfigurable Separable Convolution
Rao, L., Zhang, B., & Zhao, J. (2018) · IEEE Computer Society Annual Symposium on VLSI · doi:10.1109/isvlsi.2018.00051 · Cited by 4
“A convolutional network processor was implemented on a Virtex 4 [4]; it uses a general-purpose soft-processor for control and a 7×7 convolver for data processing.”
MethodDesign of Hardware Accelerator for Artificial Neural Networks Using Multi-operand Adder
Mayannavar, S., & Wali, U. V. (2019) · Communications in computer and information science · doi:10.1007/978-981-15-1384-8_14 · Cited by 4
A CNN accelerator on embedded FPGA using dynamic reconfigurable coprocessor
Yin, J., Yang, J., Hong-bin, Y., & Du, Z. (2019) · International Conference on Artificial Intelligence, Information Processing and Cloud Computing (AIIPCC) · doi:10.1145/3371425.3371486 · Cited by 4
Design Space Exploration of Matrix–Matrix Convolution Operation
Behera, P., & Deb, A. (2021) · Journal of Circuits Systems and Computers · doi:10.1142/s0218126621502881 · Cited by 4
FPGA-accelerated deep neural network for real-time inversion of geosteering data
Jin, Y., Wan, Q., Wu, X., Fu, X., & Chen, J. (2023) · Geoenergy Science and Engineering · doi:10.1016/j.geoen.2023.211610 · Cited by 4
A time domain 2D OaA-based convolutional neural networks accelerator
Singh, R. P., Kumar, S., Gandhi, J., Shekhawat, D., Santosh, M., & Pandey, J. G. (2023) · Memories - Materials Devices Circuits and Systems · doi:10.1016/j.memori.2023.100041 · Cited by 4
An LUT-Based Multiplier Array for Systolic Array-Based Convolutional Neural Network Accelerator
Liu, M., He, Y., & Jiao, H. (2022) · IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) · doi:10.1109/apccas55924.2022.10090287 · Cited by 4
“2 [8].”
VLSI-Friendly Filtering Algorithms for Deep Neural Networks
Cariow, A., Papliński, J. P., & Makowska, M. (2023) · Applied Sciences · doi:10.3390/app13159004 · Cited by 4
Zhang, Z., Ma, C., & Zhu, R. (2017) · Sensors · doi:10.3390/s17091941 · Cited by 4
“There are several hardware platforms for developing accelerated implementation of DNN models, including multicore CPUs [15], General Purpose Graphics Processing Units (GPGPUs) [16], Application Specific Integrated Circuits (ASICs) [17], and Field Programmable Gate Arrays (FPGAs) [18].”
MethodWang, W., Zhu, M., Chi, X., & Xu, H. (2021) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2021.102122 · Cited by 4
Lacey, G. (2016) · The Atrium (University of Guelph) · Cited by 4
Hardware Automated Datafow Deployment of CNNs
Abdelouahab, K., Pelcat, M., Sérot, J., Berry, F., Bourrasset, C., & Quinton, J. (2017) · HAL · Cited by 4
A Deep Residual Networks Accelerator on FPGA
Zhao, Y., Zhang, X., Fang, X., Li, L., Li, X., Guo, Z., & Liu, X. (2019) · International Conference on Advanced Computational Intelligence (ICACI) · doi:10.1109/icaci.2019.8778613 · Cited by 3
A 128-channel real-time VPDNN stimulation system for a visual cortical neuroprosthesis
Mohamed, H., Raducanu, B., Kiselev, I., Wang, Z., Küçükoğlu, B., Rueckauer, B., van Gerven, M. A. J., López, C. M., & Liu, S. (2023) · IEEE Biomedical Circuits and Systems Conference (BioCAS) · doi:10.1109/biocas58349.2023.10389055 · Cited by 3
“…received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 899287. such as FPGA, e.g. [7]–[10], and ASIC forms e.g. [11] are interesting for this application because they are usually more energy-efficient [12] and can incorporate…”
BackgroundMemory-Driven Data-Flow Optimization for Neural Processing Accelerators
Nie, Q. (2020) · Cited by 3
Jahiruzzaman, M., Saha, S., & Hawlader, M. A. K. (2015) · International Conference on Electrical Engineering and Information Communication Technology · doi:10.1109/iceeict.2015.7307425 · Cited by 3
“An efficient Convolution Network (ConvNets) based low and OSP oriented FPGA real-time implementation is introduced in [6]-[9].”
Visual spike-based convolution processing with a Cellular Automata architecture
Rivas-Perez, M., Linares-Barranco, A., Boluda, J. C., Ferrando, N., Jiménez, G., & Civit, A. (2010) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2010.5596924 · Cited by 3
Systolic-based 2D convolver for CNN in FPGA
Hrabovský, J., Segeč, P., Moravčík, M., & Papán, J. (2017) · International Conference on Emerging eLearning Technologies and Applications (ICETA) · doi:10.1109/iceta.2017.8102485 · Cited by 3
“[1,2] W”
InfluentialHan, W.-S., & Han, I. S. (2017) · Studies in computational intelligence · doi:10.1007/978-3-319-69266-1_13 · Cited by 3
Metadata
- Authors
- Farabet, C., Poulet, C., Han, J. Y., & LeCun, Y.
- Affiliations
- Courant Institute of Mathematical Sciences (US)
- ORCID
- LeCun, Yann
- Venue
- 2009 International Conference on Field Programmable Logic and Applications
- Publisher
- IEEE
- Published
- 2009-08
- Pages
- 32-37
- OpenAlex
- W1990315422
- Cited by
- 367
- References
- 17
- Citations by year
- 2026: 6 · 2025: 8 · 2024: 12 · 2023: 13 · 2022: 19 · 2021: 23 · 2020: 36 · 2019: 46 · 2018: 55 · 2017: 54 · 2016: 51 · 2015: 15 · 2014: 2 · 2013: 4 · 2012: 3
- FWCI
- 6.26
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- International Conference on Field Programmable Logic and Applications (FPL) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Embedded Systems and FPGA Applications
- Keywords
- convolutional neural networks, FPGA implementation, multiply-accumulate unit, face detection, embedded vision, low-power electronics, micro UAV, hardware acceleration
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.