Research map / Label-efficient and robust learning / Recognition and long-tail
Optimal Transport-based Identity Matching for Identity-invariant Facial Expression Recognition
Daeha Kim, Byung Cheol Song
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 |
|---|---|---|---|
| ELIM (AL-tuned) vs CAF (AL-tuned) | ELIM · Aff-wild | CCC (V): 0.54 -> 0.628 | Improved |
| swap backbone AL-tuned to Mlp-Mixer (ablation) | ELIM · Aff-wild | PCC (A): 0.638 -> 0.684 | Improved |
| swap backbone AL-tuned to ResNet18 (ablation) | ELIM · Aff-wild | CCC (A): 0.611 -> 0.570 | Hurt |
| swap backbone AL-tuned to Mlp-Mixer (ablation) | ELIM · Aff-wild2 validation | mean CCC: 0.465 -> 0.495 | Improved |
| ELIM R18 vs CAF R18 | ELIM · AFEW-VA static | CCC (A): 0.54 -> 0.614 | Improved |
| ELIM MMx vs ELIM R18 (ablation) | ELIM · AFEW-VA static | CCC (A): 0.614 -> 0.598 | Hurt |
| use relevance weight instead of self-side weight (Sinkhorn, no sampling) (ablation) | ELIM · AFEW-VA | mean CCC: 0.580 -> 0.587 | Improved |
| use relevance weight instead of self-side (Sinkhorn, Gumbel) (ablation) | ELIM · AFEW-VA | mean CCC: 0.583 -> 0.592 | Improved |
| add Gumbel sampling to Sinkhorn (ablation) | ELIM · AFEW-VA | elapsed time (sec): 1.97 -> 0.26 | Improved |
| add Gumbel sampling to Sinkhorn with relevance (ablation) | ELIM · AFEW-VA | mean CCC: 0.587 -> 0.592 | Improved |
| replace QPTH solver with Sinkhorn (ablation) | ELIM · AFEW-VA | elapsed time (sec): 0.97 -> 0.26 | Improved |
| replace QPTH solver with Sinkhorn (ablation) | ELIM · AFEW-VA | mean CCC: 0.586 -> 0.583 | No clear effect |
Closest papers by meaning
- Face2Exp: Combating Data Biases for Facial Expression Recognition (CVPR 2022 · similarity 0.85)
- Open-Set Facial Expression Recognition (AAAI 2024 · similarity 0.85)
- Generalizable Facial Expression Recognition (ECCV 2024 · similarity 0.85)
- QCS:Feature Refining from Quadruplet Cross Similarity for Facial Expression Recognition (AAAI 2025 · similarity 0.84)
- $\text{ID}^3$: Identity-Preserving-yet-Diversified Diffusion Models for Synthetic Face Recognition (NeurIPS 2024 · similarity 0.83)
- FAIR-FER: A Latent Alignment Approach for Mitigating Bias in Facial Expression Recognition (Student Abstract) (AAAI 2024 · similarity 0.83)
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.
