HumanOrbit: 3D Human Reconstruction as 360{\deg} Orbit Generation

TL;DR

HumanOrbit is a novel video diffusion model that generates a 360-degree view around a person from a single image, enabling high-quality 3D reconstruction with consistent appearance and identity.

Contribution

It introduces a new diffusion-based approach for multi-view human image synthesis that improves view consistency and 3D reconstruction quality.

Findings

Produces geometrically consistent multi-view images

Reconstructs textured 3D meshes with high fidelity

Outperforms state-of-the-art baselines in completeness

Abstract

We present a method for generating a full 360{\deg} orbit video around a person from a single input image. Existing methods typically adapt image-based diffusion models for multi-view synthesis, but yield inconsistent results across views and with the original identity. In contrast, recent video diffusion models have demonstrated their ability in generating photorealistic results that align well with the given prompts. Inspired by these results, we propose HumanOrbit, a video diffusion model for multi-view human image generation. Our approach enables the model to synthesize continuous camera rotations around the subject, producing geometrically consistent novel views while preserving the appearance and identity of the person. Using the generated multi-view frames, we further propose a reconstruction pipeline that recovers a textured mesh of the subject. Experimental results validate the effectiveness of HumanOrbit for multi-view image generation and that the reconstructed 3D models exhibit superior completeness and fidelity compared to those from state-of-the-art baselines.