Research map / Video / Video segmentation and tracking
Diversifying Spatial-Temporal Perception for Video Domain Generalization
Kun-Yu Lin, Jia-Run Du, Yipeng Gao, Jiaming Zhou, Wei-Shi Zheng
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 spatial grouping to the backbone (ablation) | STDN · UCF-HMDB UCF→HMDB | accuracy: 52.7 -> 54.9 | Improved |
| Add spatial grouping to the backbone (ablation) | STDN · UCF-HMDB HMDB→UCF | accuracy: 71.9 -> 73.9 | Improved |
| Add temporal dependency modeling after spatial grouping (ablation) | Backbone + SGM · UCF-HMDB UCF→HMDB | accuracy: 54.9 -> 56.7 | Improved |
| Add temporal dependency modeling after spatial grouping (ablation) | Backbone + SGM · UCF-HMDB HMDB→UCF | accuracy: 73.9 -> 75.3 | Improved |
| Use full spatial-temporal relation modeling rather than the preceding TRM variant (ablation) | Backbone + SGM · UCF-HMDB UCF→HMDB | accuracy: 54.9 -> 58.3 | Improved |
| Use full spatial-temporal relation modeling rather than the preceding TRM variant (ablation) | Backbone + SGM · UCF-HMDB HMDB→UCF | accuracy: 73.9 -> 76.2 | Improved |
| Add video-adapted MixStyle (ablation) | Backbone + STRM · UCF-HMDB UCF→HMDB | accuracy: 58.3 -> 59.3 | Improved |
| Add video-adapted MixStyle (ablation) | Backbone + STRM · UCF-HMDB HMDB→UCF | accuracy: 76.2 -> 76.6 | Improved |
| Add the full STDN feature aggregation and complete design (ablation) | Backbone + MixStyle · UCF-HMDB UCF→HMDB | accuracy: 59.3 -> 60.2 | Improved |
| Add the full STDN feature aggregation and complete design (ablation) | Backbone + MixStyle · UCF-HMDB HMDB→UCF | accuracy: 76.6 -> 77.1 | Improved |
| Add relation discrimination loss to temporal relation modeling (ablation) | STDN temporal relation module · UCF-HMDB UCF→HMDB | normalized MSE diversity | Improved |
| Use STDN instead of VideoDG | UCF-HMDB · HMDB→UCF | accuracy: 74.9 -> 77.1 | Improved |
Closest papers by meaning
- VideoRefer Suite: Advancing Spatial-Temporal Object Understanding with Video LLM (CVPR 2025 · similarity 0.87)
- VideoAnchor: Reinforcing Subspace-Structured Visual Cues for Coherent Visual-Spatial Reasoning (ICLR 2026 · similarity 0.86)
- Time-R1: Post-Training Large Vision Language Model for Temporal Video Grounding (NeurIPS 2025 · similarity 0.86)
- DELTAVID: Enhancing Fine-Grained Spatiotemporal Perception with Cross-Video Differences (NeurIPS 2026 · similarity 0.86)
- LASER: A Neuro-Symbolic Framework for Learning Spatio-Temporal Scene Graphs with Weak Supervision (ICLR 2025 · similarity 0.86)
- START: A Generalized State Space Model with Saliency-Driven Token-Aware Transformation (NeurIPS 2024 · 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.
