Research map / Video / Video generation
Video-GPT via Next Clip Diffusion
Shaobin Zhuang, Zhipeng Huang, Ying Zhang, Fangyikang Wang, Canmiao Fu, Binxin Yang, Chong Sun, Chen Li, Yali Wang
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 |
|---|---|---|---|
| increase frames per inference clip from 1 to 32 (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 0.00 -> 33.09 | Improved |
| tune history-conditioned CFG scale (3.0 vs 1) (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 25.09 -> 32.72 | Improved |
| raise CFG scale beyond 3.0 to 3.5 (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 32.72 -> 31.88 | Hurt |
| increase pretraining frames from 16 to 80 (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 22.06 -> 34.94 | Improved |
| add slight noise to clean clips during training (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 32.54 -> 33.09 | Improved |
| scale pretraining data from 1M to 70M videos (ablation) | Video-GPT · Physics-IQ | Physics-IQ score: 23.16 -> 33.09 | Improved |
| pretrain on video before fine-tuning on UCF-101 (ablation) | Video-GPT · UCF-101 class-to-video | FVD: 489 -> 53 | Improved |
| next clip diffusion pretraining vs prior models | Video-GPT · Physics-IQ | Physics-IQ score: 29.50 -> 34.97 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Seine — SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction (ICLR 2024) · better on 1 of 1 shared result
- VideoMAEv2 — VideoMAE V2: Scaling Video Masked Autoencoders With Dual Masking (CVPR 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- VideoTetris: Towards Compositional Text-to-Video Generation (NeurIPS 2024 · similarity 0.88)
- Rejuvenating image-GPT as Strong Visual Representation Learners (ICML 2024 · similarity 0.87)
- Generative Pre-trained Autoregressive Diffusion Transformer (NeurIPS 2025 · similarity 0.86)
- Towards Precise Scaling Laws for Video Diffusion Transformers (CVPR 2025 · similarity 0.86)
- Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization (ICML 2024 · similarity 0.86)
- StreamingT2V: Consistent, Dynamic, and Extendable Long Video Generation from Text (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.
