Research map / Label-efficient and robust learning / Recognition and long-tail
Face2Exp: Combating Data Biases for Facial Expression Recognition
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 |
|---|---|---|---|
| add pseudo-label training on unlabeled FR data (ablation) | Meta-Face2Exp · AffectNet (7 classes) | mean class accuracy: 60.66 -> 62.03 | Helped |
| add pseudo-label training on unlabeled FR data (ablation) | Meta-Face2Exp · RAF-DB | mean class accuracy: 84.16 -> 86.25 | Helped |
| add consistency loss with augmented FR images (ablation) | Meta-Face2Exp · AffectNet (7 classes) | mean class accuracy: 62.03 -> 63.60 | Helped |
| add consistency loss with augmented FR images (ablation) | Meta-Face2Exp · RAF-DB | mean class accuracy: 86.25 -> 87.26 | Helped |
| add circuit-feedback loss (ablation) | Meta-Face2Exp · AffectNet (7 classes) | mean class accuracy: 63.60 -> 64.23 | Helped |
| add circuit-feedback loss (ablation) | Meta-Face2Exp · RAF-DB | mean class accuracy: 87.26 -> 88.54 | Helped |
| use Meta-Face2Exp rather than supervised learning | ResNet50 · AffectNet (7 classes) | std accuracy: 21.53% to 10.07% | Helped |
| use Meta-Face2Exp rather than supervised learning | ResNet50 · RAF-DB | std accuracy: 15.48% to 9.43% | Helped |
| reduce labeled FER training data to 5% | Meta-Face2Exp · AffectNet (7 classes) | mean class accuracy: 58.37 -> 61.66 | Helped |
| reduce labeled FER training data to 50% | Meta-Face2Exp · RAF-DB | mean class accuracy: 84.16 -> 85.04 | Helped |
| increase labeled training data from 5% to 10% | Meta-Face2Exp · AffectNet (7 classes) | std accuracy: 10.69 -> 10.07 | Helped |
| increase labeled training data from 25% to 50% | Meta-Face2Exp · RAF-DB | std accuracy: 9.43 -> 10.70 | Hurt |
Closest papers by meaning
- FAIR-FER: A Latent Alignment Approach for Mitigating Bias in Facial Expression Recognition (Student Abstract) (AAAI 2024 · similarity 0.88)
- Leave No Stone Unturned: Mine Extra Knowledge for Imbalanced Facial Expression Recognition (NeurIPS 2023 · similarity 0.88)
- Open-Set Facial Expression Recognition (AAAI 2024 · similarity 0.87)
- Navigating Label Ambiguity for Facial Expression Recognition in the Wild (AAAI 2025 · similarity 0.86)
- Personalized Feature Translation for Expression Recognition: An Efficient Source-Free Domain Adaptation Method (ICLR 2026 · similarity 0.86)
- QCS:Feature Refining from Quadruplet Cross Similarity for Facial Expression Recognition (AAAI 2025 · 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.