PoseMaster: Generating 3D Characters in Arbitrary Poses from a Single Image

2025-06-26

AI Summary

Existing single-image-to-3D character generation methods suffer from self-occlusion artifacts and geometric degradation due to conventional pose normalization. To address this, we propose PoseMaster—the first end-to-end controllable 3D character generation framework. Our approach explicitly conditions generation on 3D skeletal pose, unifying pose transformation and 3D synthesis within a single architecture. We introduce a randomized masked multi-condition training strategy to improve pose control fidelity and generalization across unseen poses. Furthermore, we adopt a streaming 3D-native generation paradigm that jointly learns implicit skinning weight estimation from skeletal structure. Evaluated on a newly constructed high-fidelity pose-controlled dataset, PoseMaster achieves state-of-the-art performance in both A-pose and arbitrary-pose settings, significantly outperforming prior methods in geometric integrity and pose controllability.

Technology Category

• Computer Vision: 3D Computer Vision
• Humans and AI: Game Design — Virtual Humans, NPCs and Autonomous Characters
• Natural Language Processing: Generation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

3D characters play a crucial role in our daily entertainment. To improve the efficiency of 3D character modeling, recent image-based methods use two separate models to achieve pose standardization and 3D reconstruction of the A-pose character. However, these methods are prone to generating distorted and degraded images in the pose standardization stage due to self-occlusion and viewpoints, which further affects the geometric quality of the subsequent reconstruction process. To tackle these problems, we propose PoseMaster, an end-to-end controllable 3D character generation framework. Specifically, we unify pose transformation and 3D character generation into a flow-based 3D native generation framework. To achieve accurate arbitrary-pose control, we propose to leverage the 3D body bones existing in the skeleton of an animatable character as the pose condition. Furthermore, considering the specificity of multi-condition control, we randomly empty the pose condition and the image condition during training to improve the effectiveness and generalizability of pose control. Finally, we create a high-quality pose-control dataset derived from realistic character animation data to make the model learning the implicit relationships between skeleton and skinning weights. Extensive experiments show that PoseMaster outperforms current state-of-the-art techniques in both qualitative and quantitative evaluations for A-pose character generation while demonstrating its powerful ability to achieve precise control for arbitrary poses.

Problem

• Generating 3D characters from single images with distortions
• Improving pose standardization and 3D reconstruction quality
• Achieving precise arbitrary-pose control in character generation

Innovation

• Flow-based 3D native generation framework
• 3D body bones as pose condition
• Randomly empty conditions for better generalizability