Research map / Vision-language and multimodal / Multimodal LLMs
EMO-LLaMA: Enhancing Facial Emotion Understanding with Instruction Tuning
Bohao Xing, Zitong Yu, Xin Liu, Kaishen Yuan, Qilang Ye, Weicheng Xie, Huanjing Yue, Jingyu Yang, Heikki A. Kalviainen
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 facial embedding and ROI landmark priors (ablation) | EMO-LLaMA · MAFW | UAR: 34.75 -> 40.67 | Improved |
| add full landmark embedding (ablation) | EMO-LLaMA · MAFW | UAR: 40.67 -> 41.14 | Improved |
| add age-gender-race handcrafted prompt (ablation) | EMO-LLaMA · MAFW | UAR: 41.14 -> 41.57 | Improved |
| use Face Info Mining instead of pool and concat (ablation) | EMO-LLaMA · MAFW | UAR: 34.51 -> 40.67 | Improved |
| add category instruction data (ablation) | LLaMA-VID · AffectNet | accuracy: 32.96 -> 57.93 | Improved |
| add conversation instruction data (ablation) | EMO-LLaMA · AffectNet | accuracy: 57.93 -> 59.58 | Improved |
| add conversation instruction data (ablation) | EMO-LLaMA · MAFW | UAR: 31.56 -> 34.75 | Improved |
| add descriptive text at inference (ablation) | EMO-LLaMA · MAFW | UAR: 41.57 -> 49.29 | Improved |
| add imbalanced dataset sampling (ablation) | EMO-LLaMA · AffectNet | accuracy: 11.30 -> 57.93 | Improved |
| add imbalanced dataset sampling (ablation) | EMO-LLaMA · MAFW | WAR: 39.27 -> 44.15 | Improved |
| train on image data and test zero-shot on video versus GPT-4V | EMO-LLaMA · DFEW | UAR: 36.96 -> 44.73 | Improved |
| train on video data and test zero-shot on images versus GPT-4V | EMO-LLaMA · AffectNet | accuracy: 42.77 -> 38.43 | Hurt |
Closest papers by meaning
- Emotion-LLaMA: Multimodal Emotion Recognition and Reasoning with Instruction Tuning (NeurIPS 2024 · similarity 0.92)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.89)
- Customizing Visual Emotion Evaluation for MLLMs: An Open-vocabulary, Multifaceted, and Scalable Approach (ICLR 2026 · similarity 0.89)
- MME-Emotion: A Holistic Evaluation Benchmark for Emotional Intelligence in Multimodal Large Language Models (ICLR 2026 · similarity 0.89)
- EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 2025 · similarity 0.89)
- AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models (ICML 2025 · similarity 0.89)
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.
