Research map / Architectures and efficiency / CNN design
Low-Rank Bilinear Pooling for Fine-Grained Classification
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 |
|---|---|---|---|
| Use low-rank bilinear pooling with classifier co-decomposition | VGG16 · CUB-200-2011 | accuracy: 84.01 -> 84.21 | Helped |
| Use low-rank bilinear pooling with classifier co-decomposition | VGG16 · DTD | accuracy: 65.57 -> 65.80 | Helped |
| Use low-rank bilinear pooling with classifier co-decomposition | VGG16 · Cars | accuracy: 91.18 -> 90.92 | Hurt |
| Use low-rank bilinear pooling with classifier co-decomposition | VGG16 · Aircrafts | accuracy: 87.18 -> 87.31 | Helped |
| Replace Tensor Sketch with LRBP | VGG16 on CUB-200-2011 · CUB-200-2011 | parameter size: 8MB -> 0.8MB | Helped |
| Replace Full Bilinear with LRBP | VGG16 on CUB-200-2011 · CUB-200-2011 | parameter size: 200MB -> 0.8MB | Helped |
| Constrain classifiers to rank 10 (ablation) | Full-rank bilinear SVM classifiers · CUB-200-2011 | accuracy | No clear effect |
| Set classifier rank to r=8 and reduced dimension to m=100 (ablation) | LRBP · CUB-200-2011 | accuracy | No clear effect |
| Add ℓ2 normalization after sign square root normalization of feature maps (ablation) | LRBP-I · Fine-grained benchmarks | accuracy | Hurt |
| Use sign square root normalization on conv5_3 feature maps (ablation) | LRBP-I · Fine-grained benchmarks | accuracy | Helped |
Closest papers by meaning
- Compact Bilinear Pooling (CVPR 2016 · similarity 0.91)
- Revisiting Bilinear Pooling: A Coding Perspective (AAAI 2020 · similarity 0.90)
- Efficient Compact Bilinear Pooling via Kronecker Product (AAAI 2022 · similarity 0.90)
- Fast and Compact Bilinear Pooling by Shifted Random Maclaurin (AAAI 2021 · similarity 0.88)
- Hadamard Product for Low-rank Bilinear Pooling (ICLR 2017 · similarity 0.87)
- Rank-Transformed Dissimilarity Profiles for High-Dimensional Classification (NeurIPS 2026 · similarity 0.84)
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.