Research map / Architectures and efficiency / CNN design
Local 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 |
|---|---|---|---|
| Replace dense learnable convolution filters with fixed sparse binary filters and learnable linear combinations | CNN · MNIST | accuracy (%): 99.48 -> 99.51 | No clear effect |
| Replace dense learnable convolution filters with fixed sparse binary filters and learnable linear combinations | CNN · SVHN | accuracy (%): 95.21 -> 94.50 | Hurt |
| Replace dense learnable convolution filters with fixed sparse binary filters and learnable linear combinations | CNN · CIFAR-10 | accuracy (%): 92.95 -> 92.99 | No clear effect |
| Replace AlexNet convolutional layers with LBC layers | AlexNet · Full ImageNet | accuracy (%): 56.7821 -> 54.9454 | Hurt |
| Share the convolutional anchor weights across LBC layers (ablation) | LBCNN · CIFAR-10 | accuracy (%): 92.96 -> 92.91 | No clear effect |
| Increase output channels from q=16 to q=384 (ablation) | LBCNN · CIFAR-10 | accuracy (%): 82.74 -> 92.96 | Helped |
| Increase output channels from q=384 to q=512 (ablation) | LBCNN · CIFAR-10 | accuracy (%): 92.96 -> 92.09 | Hurt |
| Increase LBC filter size from 3×3 to 13×13 (ablation) | LBCNN · 100-class ImageNet | top-1 accuracy (%): 62.56 -> 62.43 | No clear effect |
| Use 8848 convolutional layers (4424 LBC modules) (ablation) | LBCNN · CIFAR-10 | accuracy (%): 92.99 -> 93.66 | Helped |
Closest papers by meaning
- Training binary neural networks with real-to-binary convolutions (ICLR 2020 · similarity 0.89)
- Towards Accurate Binary Convolutional Neural Network (NeurIPS 2017 · similarity 0.88)
- ClcNet: Improving the Efficiency of Convolutional Neural Network Using Channel Local Convolutions (CVPR 2018 · similarity 0.86)
- BD-Net: Has Depth-Wise Convolution Ever Been Applied in Binary Neural Networks? (AAAI 2026 · similarity 0.86)
- Compressing Deep Convolutional Neural Networks by Stacking Low-dimensional Binary Convolution Filters (AAAI 2021 · similarity 0.86)
- Role of Locality and Weight Sharing in Image-Based Tasks: A Sample Complexity Separation between CNNs, LCNs, and FCNs (ICLR 2024 · similarity 0.85)
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.