TryOnReward: Learning Foveated Consistency for Reinforcement Fine-Tuning of Virtual Try-On

2026-09-06

AI Summary

为解决虚拟试穿中评分函数与人类偏好不一致的问题，提出TryOnReward模型，通过细粒度奖励机制和视觉-语言基础优化试穿效果。

Abstract

Virtual Try-On (VTON) aims to dress a person with the reference garment, producing visually reasonable results aligned with human preferences. Turning this preference-oriented goal into an actionable objective relies on a scoring function aligned with human taste. However, classic fidelity metrics exhibit weak correlation with human judgments, and generic VLMs fail to provide the discriminative granularity demanded by try-on quality evaluation, which hinges on faithfully preserving garment and person details. This shortcoming is further exacerbated in the reinforcement fine-tuning (RFT) optimization and leads to severe reward hacking. To this end, we present TryOnReward, a fine-grained reward model tailored for VTON. Built on a vision-language backbone, it adopts a foveation calibration objective that grounds each quality dimension in the relevant region to avoid global shortcut learning. Meanwhile, TryOnReward jointly optimizes pairwise preferences and per-dimension quality scores via margin-aware supervision, leveraging both relative and absolute quality signals. For model training and evaluation, we build TryOnReward-100K, a human-annotated per-dimension rating dataset, alongside TryOn-Bench and TryOnRewardBench, two benchmarks covering diverse real scenarios. Extensive experiments confirm that TryOnReward significantly outperforms generic judges in human preference alignment, and when serving as the RFT reward function, it consistently yields human-preferred try-on results across multiple baselines.

Problem

• Virtual Try-On
• Reinforcement Fine-Tuning
• Fidelity Metrics
• Vision-Language Models
• Reward Hacking

Innovation

• Foveation Calibration
• Reinforcement Fine-Tuning (RFT)
• Pairwise Preferences
• Margin-Aware Supervision
• Vision-Language Model