Research map / Label-efficient and robust learning / Recognition and long-tail
Parametric Contrastive Learning
Jiequan Cui, Zhisheng Zhong, Shu Liu, Bei Yu, Jiaya Jia
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 |
|---|---|---|---|
| replace cross-entropy with supervised contrastive loss | ResNet-50 · ImageNet-LT | top-1 accuracy: 48.4 -> 22.0 | Hurt |
| add parametric class-wise learnable centers to supervised contrastive loss (ablation) | SupCon, ResNet-50 · ImageNet-LT | top-1 accuracy: 22.0 -> 51.0 | Improved |
| random resized crop (0.08,1) instead of (0.2,1) (ablation) | PaCo, ResNet-50 · ImageNet-LT | top-1 accuracy: 52.2 -> 55.0 | Improved |
| RandAugment on encoder and SimAugment on momentum encoder instead of SimAugment on both (ablation) | PaCo, ResNet-50 · ImageNet-LT | top-1 accuracy: 55.0 -> 57.0 | Improved |
| RandAugment on encoder with SimAugment on momentum encoder instead of RandAugment on both (ablation) | PaCo, ResNet-50 · ImageNet-LT | top-1 accuracy: 56.5 -> 57.0 | Improved |
| PaCo instead of Balanced Softmax | ResNet-50 · ImageNet-LT (RandAugment, 400 epochs) | top-1 accuracy: 55.0 -> 57.0 | Improved |
| PaCo instead of CE plus re-weighted contrastive multi-task (ablation) | ResNet-50 · ImageNet-LT | top-1 accuracy: 49.0 -> 51.0 | Improved |
| PaCo instead of multi-task Balanced Softmax plus re-weighting (loss weight 0.05) (ablation) | ResNeXt-50 · ImageNet-LT | top-1 accuracy: 57.0 -> 58.2 | Improved |
| PaCo instead of supervised contrastive learning | ResNet-50 · CIFAR-100 (full) | top-1 accuracy: 76.5 -> 79.1 | Improved |
| PaCo instead of multi-task CE plus contrastive (ablation) | ResNet-50 · ImageNet (full, RandAugment) | top-1 accuracy: 78.1 -> 79.3 | Improved |
| use RandAugment instead of SimAugment | PaCo ResNet-50 · ImageNet (full) | top-1 accuracy: 78.7 -> 79.3 | Improved |
| add RandAugment | PaCo ResNet-152 · Places-LT | top-1 accuracy: 41.2 -> 41.2 | No clear effect |
Closest papers by meaning
- Generalized Parametric Contrastive Learning (TPAMI 2023 · similarity 0.93)
- Long-Tailed Recognition by Mutual Information Maximization between Latent Features and Ground-Truth Labels (ICML 2023 · similarity 0.89)
- Contrastive Learning Based Hybrid Networks for Long-Tailed Image Classification (CVPR 2021 · similarity 0.89)
- Continuous Contrastive Learning for Long-Tailed Semi-Supervised Recognition (NeurIPS 2024 · similarity 0.89)
- Decoupled Contrastive Learning for Long-Tailed Recognition (AAAI 2024 · similarity 0.87)
- Long-tailed Object Detection Pretraining: Dynamic Rebalancing Contrastive Learning with Dual Reconstruction (NeurIPS 2024 · similarity 0.86)
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.
