Research map / Video / Video generation
MATRIX: Mask Track Alignment for Interaction-aware Video Generation
Siyoon Jin, Seongchan Kim, Jae Ho Lee, Dahyun Chung, Hyunwook Choi, Jisu Nam, Jiyoung Kim, Seungryong Kim
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 |
|---|---|---|---|
| finetune with LoRA on MATRIX-11K dataset (ablation) | CogVideoX-5B-I2V · InterGenEval ablation (MATRIX-11K finetuning) | SGI: 0.491 -> 0.526 | Improved |
| finetune with LoRA on MATRIX-11K dataset (ablation) | CogVideoX-5B-I2V · InterGenEval ablation (MATRIX-11K finetuning) | KISA: 0.406 -> 0.445 | Hurt |
| add SPA loss (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | SGI: 0.526 -> 0.540 | Improved |
| add SPA loss (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | IQ: 69.77 -> 70.26 | Improved |
| add SGA loss on t2v attention (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | KISA: 0.445 -> 0.486 | Improved |
| add SGA loss on v2t attention (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | KISA: 0.445 -> 0.509 | Improved |
| apply SGA on v2t instead of t2v attention (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | IF: 0.531 -> 0.550 | Improved |
| add SPA loss together with SGA loss (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | KISA: 0.509 -> 0.546 | Improved |
| add SPA and SGA losses together (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | IF: 0.486 -> 0.593 | Improved |
| add SPA and SGA losses together (ablation) | LoRA on MATRIX-11K · InterGenEval ablation (MATRIX-11K finetuning) | IQ: 69.77 -> 69.73 | No clear effect |
Closest papers by meaning
- SAMA: Towards Multi-Turn Referential Grounded Video Chat with Large Language Models (NeurIPS 2025 · similarity 0.88)
- HOIGen-1M: A Large-scale Dataset for Human-Object Interaction Video Generation (CVPR 2025 · similarity 0.87)
- InstanceCap: Improving Text-to-Video Generation via Instance-aware Structured Caption (CVPR 2025 · similarity 0.87)
- OmniViD: A Generative Framework for Universal Video Understanding (CVPR 2024 · similarity 0.86)
- Mask^2DiT: Dual Mask-based Diffusion Transformer for Multi-Scene Long Video Generation (CVPR 2025 · similarity 0.86)
- VideoGLaMM : A Large Multimodal Model for Pixel-Level Visual Grounding in Videos (CVPR 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.
