JVT: a simple DiT-style Transformer brings pixel-space video diffusion without VAEs
Jianjun Zhou · hf · 2026-11-16
Researchers from Zhejiang University, Shanghai Innovation Institute, and Shanghai AI Laboratory propose JVT (Just Video Transformer), a simple large-patch DiT-style Transformer for pixel-space video diffusion.
- Removes the need for pretrained autoencoders and U-Net skip connections
- Achieves competitive generation quality for class-conditioned and unconditional video
- Supports joint camera-conditioned RGB-disparity generation, enabling 3D-aware video synthesis
More from Multimodal
- Midjourney black-and-white film portrait prompt with full params — michaelrabone · 2026-11-16
- Seedance 2.5 video brings 1991 Nagasaki back to life, goes viral — SimplyAnnisa · 2026-11-16
- PixVerse demo: two mermaids having a drink in a sunken city tavern — Wise-Accident-9083 · 2026-11-16
- Blind test game suggests AI has passed the vocals Turing test too — chewygummie · 2026-11-16
- WeituAI's streaming video brain scores 69.9 on OVBench, beating Gemini baseline by 11 points — 机器之心 · 2026-11-16
- Creator crafts 4K bonfire short film frame by frame with Seedance 2.5 — gen_ericai · 2026-11-16