AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

Towards Memory- and Time-Efficient Backpropagation for Training Spiking Neural Networks

Qingyan Meng, Mingqing Xiao, Shen Yan, Yisen Wang, Zhouchen Lin, Zhi-Quan Luo

ICCV 2023 · Architectures and efficiency › Spiking networks · approach: Training deep spiking networks · 85 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
drop temporal gradient components (SLTT) instead of full BPTTSNN training, ResNet-18 · CIFAR-10training memory: 3.00G -> 1.09GImproved
drop temporal gradient components (SLTT) instead of full BPTTSNN training · CIFAR-10accuracy: 94.60% -> 94.59%No clear effect
SLTT instead of BPTTNF-ResNet-34 · ImageNettraining time: 73.8h -> 66.9hImproved
SLTT instead of BPTTVGG-11 · DVS-CIFAR10accuracy: 73.60% -> 77.30%Improved
SLTT instead of BPTTVGG-11 · DVS-Gesturetraining time: 2.68h -> 2.64hNo clear effect
backpropagate at only K=1 random time step (SLTT-1) (ablation)SLTT, NF-ResNet-34 · ImageNettraining time: 66.90h -> 32.03hImproved
backpropagate at only K=1 random time step (SLTT-1) (ablation)SLTT, NF-ResNet-34 · ImageNetaccuracy: 66.19% -> 66.17%No clear effect
backpropagate at K=4 steps (SLTT-4) instead of all 20 (ablation)SLTT, VGG-11 · DVS-Gestureaccuracy: 97.92% -> 97.45%No clear effect
SLTT instead of OTTTNF-ResNet-34 · ImageNettraining time: 7.52h -> 2.23hImproved
increase time constant tau (larger decay) (ablation)SLTT · DVS-Gesture and DVS-CIFAR10accuracyHurt
replace batch norm with weight standardization (ablation)SLTT VGG-11 · DVS-Gestureaccuracy: 97.92% -> 98.50%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy