Mesh-RFT: Enhancing Mesh Generation via Fine-grained Reinforcement Fine-Tuning

Jing Xu (Harbin Institute of Technology, Shenzhen) · Song Guo (The Hong Kong University of Science and Technology) · Jian Liu (Hong Kong University of Science and Technology) · Jing Li (The Hong Kong Polytechnic University) · Zhuo Chen (ByteDance Inc.) · jingfeng Guo (South China University of Technology) · Jiaao Yu (Beijing University of Posts and Telecommunications) · Haohan Weng (South China University of Technology) · Biwen Lei (Tencent) · Xianghui Yang (Tencent) · Fangqi Zhu (Hong Kong University of Science and Technology) · Tao Han (Shanghai AI Lab) · Chunchao Guo (Tencent Hunyuan)

3d mesh generationboundary edge ratiodata biasesfine-grained rl strategygeometric integrityglobal reinforcement learninghausdorff distancelocalized refinementmasked direct preference optimizationobject-level rewardspre-trained modelsquality-aware face maskingstate-of-the-art performancetopological regularitytopology scoretopology-aware scoring system

Existing pretrained models for 3D mesh generation often suffer from data biases and produce low-quality results, while global reinforcement learning (RL) methods rely on object-level rewards that struggle to capture local structure details. To address these challenges, we present $\textbf{Mesh-RFT}$, a novel fine-grained reinforcement fine-tuning framework that employs Masked Direct Preference Optimization (M-DPO) to enable localized refinement via quality-aware face masking. To facilitate efficient quality evaluation, we introduce an objective topology-aware scoring system to evaluate geometric integrity and topological regularity at both object and face levels through two metrics: Boundary Edge Ratio (BER) and Topology Score (TS). By integrating these metrics into a fine-grained RL strategy, Mesh-RFT becomes the first method to optimize mesh quality at the granularity of individual faces, resolving localized errors while preserving global coherence. Experiment results show that our M-DPO approach reduces Hausdorff Distance (HD) by 24.6\% and improves Topology Score (TS) by 3.8\% over pre-trained models, while outperforming global DPO methods with a 17.4\% HD reduction and 4.9\% TS gain. These results demonstrate Mesh-RFT’s ability to improve geometric integrity and topological regularity, achieving new state-of-the-art performance in production-ready mesh generation.