FPGA-BASED IMPLEMENTATION OF A SYNTHESIZABLE WEIGHT PRUNING ALGORITHM FOR DEEP NEURAL NETWORKS: AN EXPERIMENTAL APPROACH

Authors

  • Zainab A. Kadhim Department of Electronics and Communications Engineering, College of Engineering, University of Kufa, Kufa Road, Najaf, Iraq
  • Esam A. Hashim Alkaldy Department of Electronics and Communications Engineering, College of Engineering, University of Kufa, Kufa Road, Najaf, Iraq

DOI:

https://doi.org/10.30572/2018/KJE/170436

Keywords:

Machine learning, Deep neural networks, FPGA, Pruning, Accelerators

Abstract

Artificial Intelligence (AI), using deep learning techniques led to intensive computational resource consumption due to deep network with many layers and large parameters. Pruning is one compression technique that can combat these issues by reducing unnecessary weights while maintain model accuracy. In this paper, a method is proposed to pruning small weights in fully connected layers, using comparator circuits to reduce unnecessary connections in the neural network. The model was deployed on a highly parallel device platform Field Programmable Gate Array )FPGA(, the experimental result on twenty pruning configurations have shown that pruning design achieves a utilization improvement up to 6.56%, frequency enhancement of 13.8% and a power consumption reduction of 37.9% over the baseline design. The output value remains at 0.98 for the experiments without any retraining phase, thus demonstrating the effectiveness of the proposed strategy in achieving high performance and energy efficiency without incurring loss in accuracy

Downloads

Download data is not yet available.

References

Al-Sulaiman, A.M., Shubbar, R.M. and Al-Obaidi, B.H., (2025). WATER QUALITY EVALUATION OF DIWANIYAH RIVER USING THE NEMEROW POLLUTION INDEX. Kufa Journal of Engineering, 16(2), PP.1-15. https://doi.org/10.30572/2018/KJE/160201

Alsultany, A. & Karkar, A.J., (2023). Design and implementation of a parameterized elastic deep neural network accelerator with a 32-bit floating-point number using FPGA. In: Proceedings of the International Conference on Modeling & E-Information Research, Artificial Learning and Digital Applications (ICMERALDA), IEEE, pp.203–208.

Amara, A., Amiel, F. &Ea, T., (2006). FPGA vs. ASIC for low power applications. Microelectronics Journal, 37(8), pp.669–677.https://doi.org/10.1016/j.mejo.2005.11.003

Bi, F. & Yang, J., (2019). Target detection system design and FPGA implementation based on YOLO v2 algorithm. In: Proceedings of the International Conference on Imaging, Signal Processing and Communication (ICISPC), IEEE, pp.10–14.

Brown, T.B. et al., (2020). Language models are few-shot learners. Advances in Neural Information Processing Systems (NeurIPS), vol. 2020. https://doi.org/10.52202/079017-0617

Camci, E., Campolo, D.&Kayacan, E., (2020). Deep reinforcement learning for motion planning of quadrotors using raw depth images. In: Proceedings of the International Joint Conference on Neural Networks (IJCNN), IEEE, pp.1–7.

Cheng, Y., Wang, D., Zhou, P. & Zhang, T., (2017). A survey of model compression and acceleration for deep neural networks. arXiv preprint, arXiv:1710.09282.

Deng, C., Jing, D., Ding, Z. & Han, Y., (2022). Sparse channel pruning and assistant distillation for faster aerial object detection. Remote Sensing, 14(21), p.5347.

Foldy-Porto, A., Venkatesha, Y. & Panda, P., (2021). Activation density driven efficient pruning in training. In: Proceedings of the International Conference on Pattern Recognition (ICPR), IEEE, pp.8929–8936. https://doi.org/10.1109/icpr48806.2021.9413182

Fujii, T., Sato, S. & Nakahara, H., (2018). A threshold neuron pruning for a binarized deep neural network on an FPGA. IEICE Transactions on Information and Systems, 101(2), pp.376–386. https://doi.org/10.1587/transinf.2017rcp0013

Galea, T., Elsen, E. & Hooker, S., (2019). The state of sparsity in deep neural networks. arXiv preprint, arXiv:1902.09574. https://doi.org/10.20944/preprints202411.2377.v1

Garea, A.S., Heras, D.B. &Argüello, F., (2019). Caffe CNN-based classification of hyperspectral images on GPU. Journal of Supercomputing, 75, pp.1065–1077. https://doi.org/10.1007/s11227-018-2300-2

Girshick, R., Donahue, J., Darrell, T. & Malik, J., (2014). Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.580–587. https://doi.org/10.1109/cvpr.2014.81

https://doi.org/10.1109/cvpr.2014.220

https://doi.org/10.1109/icet49382.2020.9119500

https://doi.org/10.1109/icispc.2019.8935783

https://doi.org/10.1109/icmeralda60125.2023.10458208

https://doi.org/10.1109/ijcnn48605.2020.9207490

https://doi.org/10.20944/preprints202411.2377.v1

https://doi.org/10.3390/rs14215347

Huang, K. et al., (2021). Acceleration-aware fine-grained channel pruning for deep neural networks via residual gating. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 41(6), pp.1902–1915. https://doi.org/10.1109/tcad.2021.3093835

Ji, S., Xu, W., Yang, M. & Yu, K., (2013). 3D convolutional neural networks for human action recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(1), pp.221–231. https://doi.org/10.1109/tpami.2012.59

Kirchhoffer, H. et al., (2021). Overview of the neural network compression and representation (NNR) standard. IEEE Transactions on Circuits and Systems for Video Technology, 32(5), pp.3203–3216. https://doi.org/10.1109/tcsvt.2021.3095970

Liu, C., Liu, P., Zhao, W. & Tang, X., (2020). Visual tracking by structurally optimizing pre-trained CNN. IEEE Transactions on Circuits and Systems for Video Technology, 30(9), pp.3153–3166. https://doi.org/10.1109/tcsvt.2019.2938038

Marchisio, A., Hanif, M.A., Martina, M. & Shafique, M., (2018). PruNet: Class-blind pruning method for deep neural networks. In: Proceedings of the International Joint Conference on Neural Networks (IJCNN), IEEE, pp.1–8. https://doi.org/10.1109/ijcnn.2018.8489764

Muhanad, M. and Baiee, W.R., )2025(. GPS TRAJECTORY CLUSTERING FOR SPATIO–TEMPORAL BEHAVIOR ANALYSIS: THE APPLICATION OF HEATMAP TECHNIQUES AND SPATIO-TEMPORAL DUAL GRAPH NEURAL NETWORK. Kufa Journal of Engineering, 16(1), PP.57-79. https://doi.org/10.30572/2018/KJE/160105

Nane, R. et al., (2016). A survey and evaluation of FPGA high-level synthesis tools. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 35(10), pp.1591–1604. https://doi.org/10.1109/tcad.2015.2513673

See, A., Luong, M.T.& Manning, C.D., (2016). Compression of neural machine translation models via pruning. arXiv preprint, arXiv:1606.09274. https://doi.org/10.18653/v1/k16-1029

Taigman, Y., Yang, M., Ranzato, M. & Wolf, L., (2014). DeepFace: Closing the gap to human-level performance in face verification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.1701–1708.

Toshev, A. & Szegedy, C., (2014). DeepPose: Human pose estimation via deep neural networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.1653–1660. https://doi.org/10.1109/cvpr.2014.214

Verma, R. and Agarwal, M.M., (2025). ADVANCING TRAFFIC SAFETY THROUGH CAPRNN: A HYBRID CAPSULE-RECURRENT NEURAL NETWORK FOR ACCIDENT FORECASTING. Kufa Journal of Engineering, 16(4), PP.335-352. https://doi.org/10.30572/2018/KJE/160419

Wang, X., Han, Y. & Xu, Y., (2019). FPGA-based acceleration for deep convolutional neural networks: A survey. Electronics, 8(2), p.240. https://doi.org/10.1145/3316781.3317753

Zahn, O., Bustamante Jr, J., Switzer, C., Daniel, T.L. & Kutz, J.N., (2022). Pruning deep neural networks generates a sparse, bio-inspired nonlinear controller for insect flight. PLoS Computational Biology, 18(9), p.e1010512.https://doi.org/10.1371/journal.pcbi.1010512

Zhang, J. et al., (2021). A low-latency FPGA implementation for real-time object detection. In: Proceedings of the IEEE International Symposium on Circuits and Systems (ISCAS), IEEE, pp.1–5. https://doi.org/10.1109/iscas51556.2021.9401577

Zhang, S., Cao, J., Zhang, Q., Zhang, Y. & Wang, Y., (2020). An FPGA-based reconfigurable CNN accelerator for YOLO. In: Proceedings of the IEEE International Conference on Electronics Technology (ICET), IEEE, pp.74–78.

Zhao, Y., Lv, Y. & Li, C., (2023). Hardware acceleration of satellite remote sensing image object detection based on channel pruning. Applied Sciences, 13(18), p.10111. https://doi.org/10.3390/app131810111

Zhu, C. et al., (2020). An efficient hardware accelerator for structured sparse convolutional neural networks on FPGAs. IEEE Transactions on Very Large-Scale Integration (VLSI) Systems, 28(9), pp.1953–1965. https://doi.org/10.1109/tvlsi.2020.3002779

Downloads

Published

2026-11-14

How to Cite

Kadhim, Zainab A., and Esam A. Hashim Alkaldy. “FPGA-BASED IMPLEMENTATION OF A SYNTHESIZABLE WEIGHT PRUNING ALGORITHM FOR DEEP NEURAL NETWORKS: AN EXPERIMENTAL APPROACH”. Kufa Journal of Engineering, vol. 17, no. 4, Nov. 2026, pp. 695-10, https://doi.org/10.30572/2018/KJE/170436.

Share