Research map / Video / Video-language understanding
SEWA DB: A Rich Database for Audio-Visual Emotion and Sentiment Research in the Wild
Jean Kossaifi, Robert Walecki, Yannis Panagakis, Jie Shen, Maximilian Schmitt, Fabien Ringeval, Jing Han, Vedhas Pandit, Antoine Toisoul, Björn Wolfgang Schuller, Kam Star, Elnar Hajiyev and 1 more
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 |
|---|---|---|---|
| fuse landmark and SIFT features instead of landmarks alone | SVM AU detector · SEWA AU detection, test | F1 score: 0.385 -> 0.407 | Improved |
| use SVM instead of random forest | fusion features AU detector · SEWA AU detection, test | F1 score: 0.271 -> 0.407 | Improved |
| use SIFT texture features instead of landmark geometric features | SVM AU detector · SEWA AU detection, test | F1 score: 0.385 -> 0.378 | Hurt |
| optimize CCC directly instead of RMSE | ResNet-18 · SEWA valence estimation | CCC: 0.27 -> 0.35 | Improved |
| optimize CCC directly instead of RMSE | ResNet-18 · SEWA arousal estimation | CCC: 0.11 -> 0.29 | Improved |
| use video annotations instead of audio annotations for valence | SVM, AV features · SEWA valence estimation, multi-culture | CCC: 0.171 -> 0.326 | Improved |
| use audio annotations instead of video annotations for arousal | SVM, AV features · SEWA arousal estimation, multi-culture | CCC: 0.164 -> 0.267 | Improved |
| use audio-video features instead of audio features only for valence on video annotation | SVM · SEWA valence estimation, mean over cultures | CCC: 0.243681428571429 -> 0.337142857142857 | Improved |
Closest papers by meaning
- EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 2025 · similarity 0.88)
- $E^3$: Exploring Embodied Emotion Through A Large-Scale Egocentric Video Dataset (NeurIPS 2024 · similarity 0.87)
- FEEL: Quantifying Heterogeneity in Physiological Signals for Generalizable Emotion Recognition (NeurIPS 2025 · similarity 0.87)
- EEmo-Logic: A Unified Dataset and Multi-Stage Framework for Comprehensive Image-Evoked Emotion Assessment (ICML 2026 · similarity 0.86)
- OV-MER: Towards Open-Vocabulary Multimodal Emotion Recognition (ICML 2025 · similarity 0.86)
- Enriching Multimodal Sentiment Analysis Through Textual Emotional Descriptions of Visual-Audio Content (AAAI 2025 · similarity 0.86)
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.
