Research map / Architectures and efficiency / CNN design
Cross-channel Communication Networks
Jianwei Yang, Zhile Ren, Chuang Gan, Hongyuan Zhu, Devi Parikh
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 |
|---|---|---|---|
| add C3 block | ResNet-110 · CIFAR-100 | accuracy (%): 72.01 -> 73.36 | Improved |
| add C3 block | ResNet-20 · CIFAR-100 | accuracy (%): 67.73 -> 69.34 | Improved |
| add C3 block | ResNet-101 · ImageNet val | top-1 error (%): 22.48 -> 21.93 | Improved |
| add C3 block | ResNet-50 · ImageNet val | top-1 error (%): 23.61 -> 23.19 | Mixed |
| add C3 block | Faster R-CNN · COCO detection | mAP@0.5:0.95: 33.9 -> 34.8 | Improved |
| add C3 block | Deeplabv2 · PASCAL VOC 2012 segmentation | mean IoU: 75.2 -> 75.7 | Improved |
| C3 vs SE block | Wide-ResNet · CIFAR-100 | accuracy (%): 78.57 vs 78.34 | No clear effect |
| add message passing to encoder/decoder (ablation) | C3 block · CIFAR-100 | accuracy (%): 72.65 -> 73.36 | Improved |
| add encoder/decoder to message passing (ablation) | C3 block · CIFAR-100 | accuracy (%): 72.74 -> 73.36 | Improved |
| insert C3 at residual layer 2 instead of layer 1 (ablation) | ResNet-110 · CIFAR-100 | accuracy (%): 72.28 -> 72.78 | Improved |
| add C3 block to ResNet-74 to match ResNet-110 baseline (ablation) | ResNet-74 · CIFAR-100 | accuracy (%) | Improved |
Closest papers by meaning
- ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions (NeurIPS 2018 · similarity 0.85)
- Channel Equilibrium Networks for Learning Deep Representation (ICML 2020 · similarity 0.85)
- Video Classification With Channel-Separated Convolutional Networks (ICCV 2019 · similarity 0.84)
- Convolutional Channel-Wise Competitive Learning for the Forward-Forward Algorithm (AAAI 2024 · similarity 0.83)
- SparX: A Sparse Cross-Layer Connection Mechanism for Hierarchical Vision Mamba and Transformer Networks (AAAI 2025 · similarity 0.82)
- ChA-MAEViT: Unifying Channel-Aware Masked Autoencoders and Multi-Channel Vision Transformers for Improved Cross-Channel Learning (NeurIPS 2025 · 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.
