Research map / Architectures and efficiency / CNN design
Synaptic Strength For Convolutional Neural Network
CHEN LIN, Zhao Zhong, Wu Wei, Junjie Yan
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Apply Synaptic pruning, pruning 96.00% of kernels | VGG · CIFAR-10 test | error (%): 6.66 -> 6.23 | Improved |
| Apply Synaptic pruning, pruning 90.00% of kernels | ResNet-18 · CIFAR-10 test | error (%): 6.45 -> 5.06 | Improved |
| Apply Synaptic pruning, pruning 95.00% of kernels | ResNet-18 · CIFAR-10 test | error (%): 6.45 -> 5.80 | Improved |
| Apply Synaptic pruning, pruning 80.00% of kernels | DenseNet-40 · CIFAR-10 test | error (%): 5.24 -> 5.58 | Hurt |
| Apply Synaptic pruning | ResNet-50 · ImageNet validation | Top1 error (%): 24.70 -> 25.32 | Hurt |
| Apply Synaptic pruning | ResNet-50 · ImageNet validation | Top5 error (%): 7.80 -> 7.20 | Improved |
| Use Synaptic Strength instead of SSL as the connection-importance indicator | ResNet-18 · CIFAR-10 | accuracy drop | Improved |
| Include the batch-normalization scale factor in Synaptic Strength (ablation) | Synaptic pruning · CIFAR-10 | accuracy | Hurt |
| Omit kernel normalization and explicit kernel-norm parameterization (ablation) | Synaptic pruning · CIFAR-10 | accuracy | Hurt |
| Use kernel-level Synaptic pruning rather than Winograd-domain weight pruning | VGG · CIFAR-10 | error (%): 6.30 -> 7.40(+1.1); 6.20 -> 6.23(+0.03) | Improved |
Closest papers by meaning
- Pruning neural networks without any data by iteratively conserving synaptic flow (NeurIPS 2020 · similarity 0.89)
- Faster CNNs with Direct Sparse Convolutions and Guided Pruning (ICLR 2017 · similarity 0.88)
- NISP: Pruning Networks Using Neuron Importance Score Propagation (CVPR 2018 · similarity 0.87)
- A Probabilistic Approach to Neural Network Pruning (ICML 2021 · similarity 0.87)
- Importance Estimation for Neural Network Pruning (CVPR 2019 · similarity 0.87)
- Improving Deep Learning Speed and Performance Through Synaptic Neural Balance (AAAI 2025 · similarity 0.86)
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.
