WORKSHOP PAPER
Stress Testing of Spiking Neural Network-based TDC-less dToF
Jack I. MacLean1, Brian D. Stewart2, Istvan Gyongy1
1The University of Edinburgh, Institute for Integrated Micro and Nano Systems, EH9 3BF, Edinburgh, U.K.
2STMicroelectronics Imaging Division, EH3 5DA, Edinburgh, U.K.

Abstract

Spiking Neural Networks (SNNs) are naturally suited for processing discrete events, such as those produced by single-photon avalanche diode (SPAD) sensors, with the promise of lower energy consumption. This paper explores a recently proposed direct Time-of-Flight (dToF) scheme that uses an SNN with a Legendre Memory Unit (LMU) architecture for processing SPAD data. The SNN is evaluated for its performance in operating conditions beyond those considered during training, demonstrating good robustness in these new operating regimes.
Year: 2025
Workshop: IISW
URL: https://doi.org/10.60928/1tqh-z8bu
Download PDF

Keywords

Spiking Neural Networks, SNN, direct Time-of-Flight, dToF, Legendre Memory Unit, LMU, single-photon avalanche diode, SPAD, energy consumption, robustness,

References

1) Gyongy, I., Dutton, N. A., & Henderson, R. K., "Direct time-of-flight single-photon imaging", IEEE Transactions on Electron Devices, 69(6)., 2021. https://doi.org/10.1109/ted.2021.3131430
2) Taneski, F., Al Abbas, T., & Henderson, R. K., "Laser power efficiency of partial histogram direct time-of-flight LiDAR sensors", Journal of Lightwave Technology, 40(17)., 2022. https://doi.org/10.1109/jlt.2022.3187293
3) Sheehan, M. P., Tachella, J., & Davies, M. E., "Spline sketches: An efficient approach for photon counting lidar", IEEE Transactions on Computational Imaging, 10., 2024. https://doi.org/10.1109/tci.2024.3404652
4) Ingle, A., & Maier, D., "Count-free single-photon 3D imaging with race logic", IEEE Transactions on Pattern Analysis and Machine Intelligence., 2023
5) Tontini, A., Mazzucchi, S., Passerone, R., Broseghini, N., & Gasparini, L., "Histogram-less LiDAR through SPAD response linearization", IEEE Sensors Journal, 24(4)., 2023. https://doi.org/10.1109/jsen.2023.3342609
6) Milanese, T., Zhao, J., Hearn, B., & Charbon, E., "Histogram-less direct time-of-flight imaging based on a machine learning processor on FPGA", Proc. IISW 2023., 2023
7) MacLean, J. I., Stewart, B., & Gyongy, I., "TDC-less direct time-of-flight imaging using spiking neural networks", IEEE Sensors Journal, 24(20)., 2024. https://doi.org/10.1109/jsen.2024.3454974
8) Voelker, A., Kajić, I., & Eliasmith, C., "Legendre memory units: Continuous-time representation in recurrent neural networks", Advances in neural information processing systems, 32., 2019
9) Yamazaki, K., Vo-Ho, V. K., Bulsara, D., & Le, N., "Spiking neural networks and their applications: A review", Brain sciences, 12(7), 863., 2022. https://doi.org/10.3390/brainsci12070863
10) Patanwala, S. M., Gyongy, I., Mai, H., Aßmann, A., Dutton, N. A., Rae, B. R., & Henderson, R. K., "A high-throughput photon processing technique for range extension of SPAD-based LiDAR receivers", IEEE OJSSCS, 2., 2021. https://doi.org/10.1109/ojsscs.2021.3118987
11) Rasmussen, D., "NengoDL: Combining deep learning and neuromorphic modelling methods", Neuroinformatics, 17(4)., 2019. https://doi.org/10.1007/s12021-019-09424-z
12) Mushtaq, U., Akram, M. W., Prasad, D., & Islam, A., "An energy and area-efficient spike frequency adaptable LIF neuron for spiking neural networks", Computers and Electrical Engineering, 119., 2024. https://doi.org/10.1016/j.compeleceng.2024.109562