# Back to Pixels: Lifting Latent Video Diffusion into Pixel Space

**Authors:** Yanke Song, Shuo Chen, Chris Rockwell, Guande He, Nan Liu, Desai Xie, Tianpei Gu, Yong-Hyun Park, Aryan V S, Dennis Yuren Menn, Ming-Hsuan Yang, Xun Huang  
**Lab:** Roblox Corporation · **Published:** 12 Nov 2026 · **Upvotes:** 3  
**Links:** [arXiv](https://arxiv.org/abs/2611.17336) · [PDF](https://arxiv.org/pdf/2611.17336) · [This page](https://omen-bit.github.io/ai-paper-daily/paper/2611.17336.html)

## Key numbers

- **2×**: it generates at 2× higher resolution with the same per-frame token count

## Abstract

Frontier video generation models work in a VAE latent space that compresses video in both space and time. This compression made training tractable, but now limits high-resolution, controllable, and interactive generation -- doubling resolution quadruples token count, conditions apply to groups of frames rather than individual ones, and each response must wait for the rest of its multi-frame block. We present PixVid, a simple yet effective method that lifts a pretrained latent video model to directly generate every frame in pixel space. PixVid keeps the diffusion transformer and replaces the VAE decoder with a convolutional head that renders its outputs into pixels. We find that predicting clean video rather than velocity, together with a larger noise shift, is key to quality. Finetuned from Wan2.1-14B, PixVid is competitive with the latent baseline within ~ 1.3M training clips. Working framewise in pixels, it generates at 2× higher resolution with the same per-frame token count, keeps depth-conditioned video editing aligned under fast motion, and responds to each action at the next frame. Project page: https://chenshuo20.github.io/PixVid/

---

Notes written by AI from the paper, numbers copied from the abstract. Check the paper before you rely on them. From AI Papers Daily.