Research map / Architectures and efficiency / CNN design
Learning Channel-Wise Interactions for Binary Convolutional Neural Networks
Check 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 |
|---|---|---|---|
| add channel-wise interactions through interacted bitcount (ablation) | CI-BCNN · MNIST | consistent-sign pixel ratio: 0.6238 -> 0.6638 | Helped |
| add channel-wise interactions through interacted bitcount (ablation) | CI-BCNN · MNIST | consistent-sign pixel ratio: 0.6061 -> 0.6244 | Helped |
| add channel-wise interactions through interacted bitcount (ablation) | CI-BCNN · MNIST | accuracy: 99.01 -> 99.10 | Helped |
| add extra shortcuts (ablation) | CI-BCNN · ImageNet | top-1 accuracy: 56.73 -> 59.90; 62.41 -> 64.93 | Helped |
| add extra shortcuts (ablation) | CI-BCNN · ImageNet | top-5 accuracy: 80.12 -> 84.18; 84.35 -> 86.61 | Helped |
| use CI-BCNN instead of Bi-Real-Net | binary ResNet18 · ImageNet | top-1 accuracy: 56.40 -> 56.73 | Helped |
| use CI-BCNN instead of Bi-Real-Net | binary ResNet18 · ImageNet | top-5 accuracy: 79.50 -> 80.12 | Helped |
| use CI-BCNN instead of Bi-Real-Net | binary ResNet34 · ImageNet | top-1 accuracy: 62.20 -> 62.41 | Helped |
| use CI-BCNN instead of Bi-Real-Net | binary ResNet34 · ImageNet | top-5 accuracy: 83.90 -> 84.35 | Helped |
| use CI-BCNN instead of LQ-Net | VGG-small · CIFAR-10 | accuracy: 93.40 -> 92.47 | Hurt |
| use CI-BCNN instead of TTQ | ResNet20 · CIFAR-10 | accuracy: 91.13 -> 91.10 | Hurt |
| increase the maximal density of channel-interaction edges (ablation) | CI-BCNN ResNet18 · ImageNet | top-1 accuracy | Mixed |
Closest papers by meaning
- ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions (NeurIPS 2018 · similarity 0.86)
- Channel Interaction Networks for Fine-Grained Image Categorization (AAAI 2020 · similarity 0.86)
- BD-Net: Has Depth-Wise Convolution Ever Been Applied in Binary Neural Networks? (AAAI 2026 · similarity 0.85)
- Monitoring Primitive Interactions During the Training of DNNs (AAAI 2025 · similarity 0.83)
- Convolutional Channel-Wise Competitive Learning for the Forward-Forward Algorithm (AAAI 2024 · similarity 0.82)
- Layerwise Change of Knowledge in Neural Networks (ICML 2024 · similarity 0.82)
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.