Research map / Architectures and efficiency / CNN design
Deep Global Generalized Gaussian 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 Gaussian covariance pooling with generalized-Gaussian covariance pooling | ResNet-50 · ImageNet-1K | top-1 error: 21.95 -> 21.31 | Helped |
| Replace Gaussian covariance pooling with generalized-Gaussian covariance pooling | ResNet-50 · Places365 | top-1 error: 43.68 -> 43.07 | Helped |
| Add robust covariance estimation (ablation) | 3G-Net with ResNet-50 · ImageNet-1K | top-1 error: 25.17 -> 21.31 | Helped |
| Add robust covariance estimation (ablation) | 3G-Net with ResNet-101 · ImageNet-1K | top-1 error: 24.23 -> 20.37 | Helped |
| Use three re-weighting iterations instead of one (ablation) | 3G-Net with ResNet-50 · ImageNet-1K | top-1 error: 21.64 -> 21.31 | Helped |
| Replace the full re-weighted block with URM-v1 (ablation) | 3G-Net with ResNet-50 · ImageNet-1K | top-1 error: 21.31 -> 21.97 | Hurt |
| Replace the full re-weighted block with URM-v2 (ablation) | 3G-Net with ResNet-50 · ImageNet-1K | top-1 error: 21.31 -> 21.89 | Hurt |
| Add the element-wise power operation to the re-weighted block (ablation) | 3G-Net · ImageNet-1K | classification performance | No clear effect |
Closest papers by meaning
- What Deep CNNs Benefit From Global Covariance Pooling: An Optimization Perspective (CVPR 2020 · similarity 0.90)
- G2DeNet: Global Gaussian Distribution Embedding Network and Its Application to Visual Recognition (CVPR 2017 · similarity 0.88)
- Global Second-Order Pooling Convolutional Networks (CVPR 2019 · similarity 0.88)
- Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization (CVPR 2018 · similarity 0.87)
- Exact, Tractable Gauss-Newton Optimization in Deep Reversible Architectures Reveal Poor Generalization (NeurIPS 2024 · similarity 0.84)
- Variational Partial Group Convolutions for Input-Aware Partial Equivariance of Rotations and Color-Shifts (ICML 2024 · similarity 0.83)
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.