| 講演抄録/キーワード |
| 講演名 |
2022-01-24 17:10
Ternarizing Deep Spiking Neural Network ○Man Wu・Yirong Kan・Van_Tinh Nguyen・Renyuan Zhang・Yasuhiko Nakashima(NAIST) VLD2021-61 CPSY2021-30 RECONF2021-69 |
| 抄録 |
(和) |
The feasibility of ternarizing spiking neural networks (SNNs) is studied in this work toward trading a slight accuracy for significantly reducing computational and memory costs. By leveraging a parametric integrate-and-fire (PIF) neuron with learnable threshold and spike-timing-dependent backpropagation (STDB) learning rule, the ternarized spiking neural networks (TSNNs) enable directly trained with low latency and negligible loss of accuracy. To this end, a paradigm for binary-ternary dotproduct operation is realized during the inference; therefore, the TSNNs achieve up to 16x model compression in contrast to the full precision SNNs. Moreover, to mitigate the accuracy gap, an optimized TSNN with a spiking ResNet structure is introduced into TSNN. For proof-of-concept, we evaluate the prototype of proposed TSNN on N-MNIST, CIFAR-10, CIFAR-100, which achieve 98.43%, 89.07%, 65.24% accuracy with 4 timesteps, respectively. On the basis of this prototype, the optimized TSNN improves by 0.84% and 0.51% over CIFAR-10 and CIFAR-100 datasets, respectively. |
| (英) |
The feasibility of ternarizing spiking neural networks (SNNs) is studied in this work toward trading a slight accuracy for significantly reducing computational and memory costs. By leveraging a parametric integrate-and-fire (PIF) neuron with learnable threshold and spike-timing-dependent backpropagation (STDB) learning rule, the ternarized spiking neural networks (TSNNs) enable directly trained with low latency and negligible loss of accuracy. To this end, a paradigm for binary-ternary dotproduct operation is realized during the inference; therefore, the TSNNs achieve up to 16x model compression in contrast to the full precision SNNs. Moreover, to mitigate the accuracy gap, an optimized TSNN with a spiking ResNet structure is introduced into TSNN. For proof-of-concept, we evaluate the prototype of proposed TSNN on N-MNIST, CIFAR-10, CIFAR-100, which achieve 98.43%, 89.07%, 65.24% accuracy with 4 timesteps, respectively. On the basis of this prototype, the optimized TSNN improves by 0.84% and 0.51% over CIFAR-10 and CIFAR-100 datasets, respectively. |
| キーワード |
(和) |
deep spiking neural network / ternary weights / SNN compression / TSNN / / / / |
| (英) |
deep spiking neural network / ternary weights / SNN compression / TSNN / / / / |
| 文献情報 |
信学技報, vol. 121, no. 343, CPSY2021-30, pp. 67-72, 2022年1月. |
| 資料番号 |
CPSY2021-30 |
| 発行日 |
2022-01-17 (VLD, CPSY, RECONF) |
| ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
| PDFダウンロード |
VLD2021-61 CPSY2021-30 RECONF2021-69 |
|