Abstract
SNN (Spiking Neural Network) is well suited for DVS (Dynamic Vision Sensor) object recognition because the output of the DVS sensor is the spike. The existing SNNs usually build networks by fully connection with a large number of parameters. However, the deep network is unable to train with this connection and large parameter networks cannot be deployed where storage is limited. To overcome these shortcomings, we introduce a new model called Fire module. There are two structures in Fire module. One is a combination of weight sharing layers and the other is a skip connection, which reduces the number of parameters and makes deep network trainable respectively. We compare our method with existing SNNs and show that our method achieves competitive performance with 1800x fewer parameters against fully connection on TMV3-DVS and N-CARS datasets. Moreover, we combine the DVS sensor and our lightweight SNN object recognition network to produce an object recognition hardware system.
This work was partially supported by the NSF of China (No. 62022063, No. 61772388 and No. 61632019).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bauer, F.C., Muir, D.R., Indiveri, G.: Real-time ultra-low power ECG anomaly detection using an event-driven neuromorphic processor. IEEE Trans. Biomed. Circuits Syst. 13(6), 1575–1582 (2019). https://doi.org/10.1109/TBCAS.2019.2953001
Cao, Y., Chen, Y., Khosla, D.: Spiking deep convolutional neural networks for energy-efficient object recognition. Int. J. Comput. Vision 113(1), 54–66 (2014). https://doi.org/10.1007/s11263-014-0788-3
Delbrück, T., Linares-Barranco, B., Culurciello, E., Posch, C.: Activity-driven, event-based vision sensors. In: Proceedings of 2010 IEEE International Symposium on Circuits and Systems, pp. 2426–2429 (2010). https://doi.org/10.1109/ISCAS.2010.5537149
Diehl, P.U., Neil, D., Binas, J., Cook, M., Liu, S., Pfeiffer, M.: Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing. In: 2015 International Joint Conference on Neural Networks (IJCNN), pp. 1–8 (2015). https://doi.org/10.1109/IJCNN.2015.7280696
Gerstner, W.: Time structure of the activity in neural network models. Phys. Rev. E 51(1), 738–758 (1995). https://doi.org/10.1103/PhysRevE.51.738
Gütig, R., Sompolinsky, H.: The tempotron: a neuron that learns spike timing–based decisions. Nat. Neurosci. 9(3), 420–428 (2006). https://doi.org/10.1038/nn1643
Lee, J.H., Delbruck, T., Pfeiffer, M.: Training deep spiking neural networks using backpropagation. Front. Neurosci. 10, 508 (2016). https://doi.org/10.3389/fnins.2016.00508
Mostafa, H.: Supervised learning based on temporal coding in spiking neural networks. IEEE Trans. Neural Netw. Learn. Syst. 29(7), 3227–3235 (2018). https://doi.org/10.1109/TNNLS.2017.2726060
Orchard, G., Meyer, C., Etienne-Cummings, R., Posch, C., Thakor, N., Benosman, R.: HFirst: a temporal approach to object recognition. IEEE Trans. Pattern Anal. Mach. Intell. 37(10), 2028–2040 (2015). https://doi.org/10.1109/TPAMI.2015.2392947
Panda, P., Roy, K.: Unsupervised regenerative learning of hierarchical features in Spiking Deep Networks for object recognition. In: 2016 International Joint Conference on Neural Networks (IJCNN), pp. 299–306 (2016). https://doi.org/10.1109/IJCNN.2016.7727212
Pei, J., et al.: Towards artificial general intelligence with hybrid Tianjic chip architecture. Nature 572(7767), 106–111 (2019). https://doi.org/10.1038/s41586-019-1424-8
Ponulak, F., Kasiński, A.: Supervised learning in spiking neural networks with ReSuMe: sequence learning, classification, and spike shifting. Neural Comput. 22(2), 467–510 (2010). https://doi.org/10.1162/neco.2009.11-08-901
Posch, C., Matolin, D., Wohlgenannt, R.: A QVGA 143 dB dynamic range frame-free PWM image sensor with lossless pixel-level video compression and time-domain CDS. IEEE J. Solid-State Circuits 46(1), 259–275 (2011). https://doi.org/10.1109/JSSC.2010.2085952
Rueckauer, B., Lungu, I.A., Hu, Y., Pfeiffer, M., Liu, S.C.: Conversion of continuous-valued deep networks to efficient event-driven networks for image classification. Front. Neurosci. 11, 682 (2017). https://doi.org/10.3389/fnins.2017.00682
Shrestha, S.B., Orchard, G.: SLAYER: spike layer error reassignment in time. arXiv:1810.08646 [cs, stat] (2018)
Sironi, A., Brambilla, M., Bourdis, N., Lagorce, X., Benosman, R.: HATS: histograms of averaged time surfaces for robust event-based object classification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1731–1740 (2018)
Zeng, Y., Zhang, T., Xu, B.: Improving multi-layer spiking neural networks by incorporating brain-inspired rules. Sci. China Inf. Sci. 60(5), 1–11 (2017). https://doi.org/10.1007/s11432-016-0439-4
Zhang, T., Zeng, Y., Zhao, D., Shi, M.: A plasticity-centric approach to train the non-differential spiking neural networks. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 32(1) (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Liu, Z., Huang, B., Wu, J., Shi, G. (2021). Lightweight Convolutional SNN for Address Event Representation Signal Recognition. In: Fang, L., Chen, Y., Zhai, G., Wang, J., Wang, R., Dong, W. (eds) Artificial Intelligence. CICAI 2021. Lecture Notes in Computer Science(), vol 13069. Springer, Cham. https://doi.org/10.1007/978-3-030-93046-2_26
Download citation
DOI: https://doi.org/10.1007/978-3-030-93046-2_26
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-93045-5
Online ISBN: 978-3-030-93046-2
eBook Packages: Computer ScienceComputer Science (R0)