Research map / Label-efficient and robust learning / Semi-supervised learning
CoSSL: Co-Learning of Representation and Classifier for Imbalanced Semi-Supervised Learning
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 CoSSL with FixMatch rather than the strongest listed baseline, DARP + cRT | FixMatch · CIFAR-10-LT, γ=150, uniform test | accuracy: 79.6 -> 80.3 | Helped |
| Use CoSSL with ReMixMatch rather than DARP + cRT | ReMixMatch · CIFAR-10-LT, γ=150, uniform test | accuracy: 79.7 -> 81.3 | Helped |
| Use CoSSL rather than DARP + cRT | ReMixMatch · CIFAR-100-LT, γ=100, uniform test | accuracy: 43.7 -> 44.1 | Helped |
| Use CoSSL rather than the strongest listed baseline | FixMatch · Small-ImageNet-127, 32×32 | averaged class recall: 40.9 -> 43.7 | Helped |
| Use CoSSL rather than the strongest listed baseline | FixMatch · Small-ImageNet-127, 64×64 | averaged class recall: 55.9 -> 53.8 | Hurt |
| Use CoSSL rather than CReST+ + LA | FixMatch · Food-101-LT, γ=100 | averaged class recall: 36.1 -> 40.4 | Helped |
| Use CoSSL rather than CReST+ | FixMatch · CIFAR-10-LT, γ=150, unknown shifted test, mean | accuracy: 71.98 -> 81.06 | Helped |
| Use CoSSL with post-compensation rather than CReST+ with post-compensation | FixMatch · CIFAR-10-LT, γ=150, known shifted test, mean | accuracy: 82.62 -> 86.20 | Helped |
| Allow gradients from classifier learning to update the representation (ablation) | CoSSL · CIFAR-10-LT, γ=150 | accuracy: 80.24 -> 76.46 | Hurt |
| Generate pseudo-labels with h_CL rather than h_SSL (ablation) | CoSSL · CIFAR-10-LT, γ=150 | accuracy: 78.23 -> 80.24 | Helped |
| Use co-learning instead of two-stage training (ablation) | CoSSL · CIFAR-10-LT, γ=150 | accuracy: 73.52 -> 80.24 | Helped |
| Add TFE instead of no classifier enhancement (ablation) | CoSSL · CIFAR-10-LT, γ=150 | accuracy: 77.22 -> 80.24 | Helped |
Closest papers by meaning
- BaCon: Boosting Imbalanced Semi-supervised Learning via Balanced Feature-Level Contrastive Learning (AAAI 2024 · similarity 0.90)
- CDMAD: Class-Distribution-Mismatch-Aware Debiasing for Class-Imbalanced Semi-Supervised Learning (CVPR 2024 · similarity 0.88)
- LCGC: Learning from Consistency Gradient Conflicting for Class-Imbalanced Semi-Supervised Debiasing (AAAI 2025 · similarity 0.88)
- Twice Class Bias Correction for Imbalanced Semi-supervised Learning (AAAI 2024 · similarity 0.87)
- BEM: Balanced and Entropy-based Mix for Long-Tailed Semi-Supervised Learning (CVPR 2024 · similarity 0.87)
- Semi-supervised Multi-label Learning with Balanced Binary Angular Margin Loss (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.