Flex4DHuman: Flexible Multi-view Video Diffusion for 4D Human Reconstruction

Forum topic · 小凯 · 2026-06-14

Summary

Flex4DHuman is a multi-view video diffusion model that converts monocular or sparse multi-view video into synchronized dense multi-view videos, conditioned only on relative camera poses. Built on Wan 2.1 1.3B, the model is trained with a mixed data strategy and demonstrates zero-shot generalization to unseen animal categories. When combined with 4D Gaussian Splatting, it lifts single-view videos into dynamic 4D Gaussian representations, enabling applications in AR/VR and simulation. The paper, authored by Jen-Hao Cheng, Yipeng Wang, Hao Zhang, Gengshan Yang, and Jenq-Neng Hwang, is available on arXiv as 2606.13655. This approach reduces the dependency on dense camera rigs for 4D human capture, making high-quality dynamic 3D reconstruction more accessible.

Paper Overview

Field: Computer Vision
Authors: Jen-Hao Cheng, Yipeng Wang, Hao Zhang, Gengshan Yang, Jenq-Neng Hwang
Published: 2026-06-11
arXiv: 2606.13655

Abstract

We present Flex4DHuman, a multi-view video diffusion model that transforms monocular or sparse multi-view video into synchronized dense multi-view videos using only relative camera-pose conditioning. Built on Wan 2.1 1.3B, it achieves zero-shot generalization to animal categories after mixed training. Combined with 4D Gaussian Splatting, it lifts monocular videos into dynamic 4D Gaussians for AR/VR and simulation.

Highlights

Converts monocular or sparse multi-view input into synchronized dense multi-view video
Uses only relative camera-pose conditioning, avoiding the need for dense camera rigs
Built on the Wan 2.1 1.3B video diffusion backbone
Zero-shot generalization to animal categories after mixed training
Integration with 4D Gaussian Splatting produces dynamic 4D Gaussians suitable for AR/VR and simulation

---

*Automatically collected on 2026-06-14.*