Research map / Label-efficient and robust learning / Recognition and long-tail
Leave No Stone Unturned: Mine Extra Knowledge for Imbalanced Facial Expression Recognition
Read the paperCheck 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 re-balanced attention consistency (RAC) to baseline (ablation) | ResNet-18 · RAF-DB | mean accuracy: 78.53 -> 80.54 | Helped |
| add re-balanced smooth labels (RSL) to baseline (ablation) | ResNet-18 · RAF-DB | mean accuracy: 78.53 -> 79.35 | Helped |
| add RSL on top of RAC (ablation) | RAC · RAF-DB | mean accuracy: 80.54 -> 82.44 | Helped |
| add RAC on top of RSL (ablation) | RSL · RAF-DB | mean accuracy: 79.35 -> 82.44 | Helped |
| use intensity transformation instead of flip for consistency (ablation) | RAC · RAF-DB | mean accuracy: 82.44 -> 77.03 | Hurt |
| use scaling instead of flip for consistency (ablation) | RAC · RAF-DB | mean accuracy: 82.44 -> 82.19 | No clear effect |
| apply method to different backbones | MobileNet · RAF-DB | mean accuracy: 74.34 -> 77.24 | Helped |
| apply method to Swin-T backbone | Swin-T · RAF-DB | mean accuracy: 84.68 -> 87.71 | Helped |
| vary consistency weight lambda 0.05 to 4 (ablation) | Ours · RAF-DB | mean accuracy | No clear effect |
Closest papers by meaning
- Navigating Label Ambiguity for Facial Expression Recognition in the Wild (AAAI 2025 · similarity 0.90)
- Open-Set Facial Expression Recognition (AAAI 2024 · similarity 0.88)
- When Facial Expression Recognition Meets Few-Shot Learning: A Joint and Alternate Learning Framework (AAAI 2022 · similarity 0.88)
- FAIR-FER: A Latent Alignment Approach for Mitigating Bias in Facial Expression Recognition (Student Abstract) (AAAI 2024 · similarity 0.88)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.87)
- EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 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.