AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

Towards Ultra Low Latency Spiking Neural Networks for Vision and Sequential Tasks Using Temporal Pruning

Sayeed Shafayet Chowdhury, Nitin Rathi, Kaushik Roy

ECCV 2022 · Architectures and efficiency › Spiking networks · approach: Training deep spiking networks · 38 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
add gradual temporal pruning instead of direct conversion to T1 (ablation)SNN VGG6 · CIFAR10top-1 accuracy (%): 88.64 -> 89.10Improved
add gradual temporal pruning instead of direct T5 to T1 transition (ablation)SNN VGG16 · CIFAR10training convergenceImproved
train ANN with batch-norm before conversion (ablation)SNN T1 VGG16 · CIFAR10top-1 accuracy (%): 91.03 -> 93.05Improved
train ANN with batch-norm before conversion (ablation)SNN T1 VGG16 · CIFAR100top-1 accuracy (%): 67.76 -> 70.15Improved
train ANN with batch-norm before conversion (ablation)SNN T5 VGG16 · CIFAR10top-1 accuracy (%): 92.15 -> 93.90Improved
include all intermediate timesteps rather than skipping to T3 (ablation)SNN T1 VGG16 · CIFAR10top-1 accuracy (%): 93.01% and 93.05%No clear effect
include all intermediate timesteps rather than skipping to T3 (ablation)SNN T1 VGG16 · CIFAR100top-1 accuracy (%): 69.92% and 70.15%No clear effect
include all intermediate timesteps rather than skipping to T3 (ablation)SNN T1 VGG16 · ImageNettop-1 accuracy (%): 68.98% and 69%No clear effect
split fixed epochs equally across pruning stages (early stop parents) (ablation)SNN T1 VGG16 · CIFAR10top-1 accuracy (%): < 0.3%No clear effect
apply magnitude weight pruning keeping 10% of weightsSNN T1 VGG16 · CIFAR10top-1 accuracy (%): 93.05 -> 91.15Hurt
apply magnitude weight pruning keeping 10% of weightsSNN T1 VGG16 · CIFAR100top-1 accuracy (%): 70.15 -> 68.20Hurt
increase timesteps from 1 to 3 (ablation)SNN DQN · Cartpolemean reward: 38.7 -> 52.2Improved

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