Research map / Architectures and efficiency / Spiking networks
Deep Residual Learning in Spiking Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace standard spiking residual blocks with SEW ADD blocks | Spiking ResNet on ImageNet · ImageNet validation | Acc@1 (%): 61.86 -> 67.04 | Helped |
| increase depth from ResNet-18 to ResNet-34 | SEW ResNet (ADD) · ImageNet validation | Acc@1 (%): 63.18 -> 67.04 | Helped |
| increase depth from ResNet-18 to ResNet-34 | Spiking ResNet · ImageNet validation | Acc@1 (%): 62.32 -> 61.86 | Hurt |
| replace standard spiking residual blocks with SEW ADD blocks | Spiking ResNet on DVS Gesture · DVS Gesture | accuracy (%): 90.97% v.s. 97.92% | Helped |
| use IAND instead of ADD (ablation) | 7B-Net SEW blocks · DVS Gesture | accuracy (%): 97.92 -> 95.49 | Hurt |
| use AND instead of ADD (ablation) | 7B-Net SEW blocks · DVS Gesture | accuracy (%): 97.92 -> 70.49 | Hurt |
| remove the shortcut connection (ablation) | 7B-Net · DVS Gesture | accuracy (%): 97.92 -> 91.67 | Hurt |
| replace ArcTan surrogate with Rectangular (ablation) | SEW ResNet (ADD) · CIFAR-10 | test accuracy: 0.8263 -> 0.8256 | No clear effect |
| replace ArcTan surrogate with Constant 1 (ablation) | SEW ResNet (ADD) · CIFAR-10 | test accuracy: 0.8263 -> 0.1256 | Hurt |
| replace ArcTan surrogate with Rectangular (ablation) | Spiking ResNet · CIFAR-10 | test accuracy: 0.7733 -> 0.6601 | Hurt |
| apply random temporal delete during training (ablation) | Plain Net, Spiking ResNet, and SEW ResNet · DVS Gesture | test accuracy | Helped |
| use IAND instead of AND (ablation) | SEW ResNet · DVS Gesture | accuracy (%): 70.49% v.s. 95.49% | Helped |
Closest papers by meaning
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (NeurIPS 2024 · similarity 0.89)
- Take A Shortcut Back: Mitigating the Gradient Vanishing for Training Spiking Neural Networks (NeurIPS 2024 · similarity 0.88)
- Efficient ANN-Guided Distillation: Aligning Rate-based Features of Spiking Neural Networks through Hybrid Block-wise Replacement (CVPR 2025 · similarity 0.88)
- Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation (NeurIPS 2024 · similarity 0.87)
- Enhanced Self-Distillation Framework for Efficient Spiking Neural Network Training (NeurIPS 2025 · similarity 0.87)
- Memory-Efficient Reversible Spiking Neural Networks (AAAI 2024 · similarity 0.87)
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.