Research map / Video / Video segmentation and tracking
Weakly Supervised Video Emotion Detection and Prediction via Cross-Modal Temporal Erasing Network
Zhicheng Zhang, Lijuan Wang, Jufeng Yang
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 intra-modal attention (temporal non-local correlation) (ablation) | visual stream · VideoEmotion-8 and Ekman-6 | accuracy: +1.3 (Video-8), +1.7 (Ekman-6) | Improved |
| add temporal erasing module on top of intra-modal attention (ablation) | visual stream · VideoEmotion-8 and Ekman-6 | accuracy: +0.5 (Video-8), +1.6 (Ekman-6) | Improved |
| add inter-modal attention and intra-modal attention with erasing (full model) (ablation) | visual+audio model · VideoEmotion-8 and Ekman-6 | accuracy: Full 57.3% vs 55.2% (intra+erasing) on Video-8; 58.2% vs 56.1% on Ekman-6 | Improved |
| temporal erasing module applied to audio-only stream (ablation) | audio stream · VideoEmotion-8 and Ekman-6 | accuracy: +1.5 (Video-8), +3.6 (Ekman-6) | Improved |
| proposed cross-modal temporal erasing network vs VAANet | visual+audio model · VideoEmotion-8, Ekman-6, CAER | accuracy: +2.8 (Video-8), +2.9 (Ekman-6), +1.8 (CAER) | Improved |
| threshold theta for erasing mask set high (above 0.7) (ablation) | temporal erasing module · Ekman-6 and VideoEmotion-8 | accuracy | Hurt |
Closest papers by meaning
- CoLeaF: A Contrastive-Collaborative Learning Framework for Weakly Supervised Audio-Visual Video Parsing (ECCV 2024 · similarity 0.87)
- Local-Global Multi-Modal Distillation for Weakly-Supervised Temporal Video Grounding (AAAI 2024 · similarity 0.86)
- Joint-Modal Label Denoising for Weakly-Supervised Audio-Visual Video Parsing (ECCV 2022 · similarity 0.86)
- An End-to-End Visual-Audio Attention Network for Emotion Recognition in User-Generated Videos (AAAI 2020 · similarity 0.86)
- MART: Masked Affective RepresenTation Learning via Masked Temporal Distribution Distillation (CVPR 2024 · similarity 0.86)
- $E^3$: Exploring Embodied Emotion Through A Large-Scale Egocentric Video Dataset (NeurIPS 2024 · similarity 0.84)
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.
