Research map / Label-efficient and robust learning / Recognition and long-tail
Weakly-Supervised Text-Driven Contrastive Learning for Facial Behavior Understanding
Xiang Zhang, Taoyue Wang, Xiaotian Li, Huiyuan Yang, Lijun Yin
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 |
|---|---|---|---|
| add activity-text weakly-supervised pre-training on top of image-only pre-training with label names and descriptions (ablation) | CLEF · BP4D | average F1 (%): 64.8 -> 65.9 | Improved |
| add activity-text pre-training (ablation) | CLEF · RAF-DB | accuracy (%): 89.44 -> 90.09 | Improved |
| add label names to CLIP image encoder (ablation) | CLIP · BP4D | average F1 (%): 63.4 -> 64.0 | Improved |
| add label descriptions in addition to label names (ablation) | CLIP · RAF-DB | accuracy (%): 88.72 -> 89.70 | Improved |
| add activity-text pre-training without label text (ablation) | CLEF · BP4D | average F1 (%): 65.0 -> 64.2 | Hurt |
| add label names only after image pre-training (ablation) | CLEF · RAF-DB | accuracy (%): 89.67 -> 88.57 | Hurt |
| use a shared text encoder for label names and descriptions instead of two (ablation) | CLEF · BP4D | average F1 (%): 64.5 -> 65.9 | Improved |
| apply CLEF pretraining and text-driven fine-tuning vs FaRL in zero-shot | FaRL · RAF-DB zero-shot | accuracy (%): 16.21 -> 29.14 | Improved |
| CLEF versus DMUE on AffectNet-8 | CLEF · AffectNet-8 | accuracy (%): 62.84 -> 62.77 | No clear effect |
Closest papers by meaning
- A Compact Embedding for Facial Expression Similarity (CVPR 2019 · similarity 0.85)
- Emotion-Aware Multi-View Contrastive Learning for Facial Emotion Recognition (ECCV 2022 · similarity 0.85)
- From Pixels to Semantics: Unified Facial Action Representation Learning for Micro-Expression Analysis (ICLR 2026 · similarity 0.84)
- Contrastive Decoupled Representation Learning and Regularization for Speech-Preserving Facial Expression Manipulation (IJCV 2025 · similarity 0.84)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.83)
- Weakly-Supervised Contrastive Learning for Imprecise Class Labels (ICML 2025 · 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.
