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

arXiv: https://arxiv.org/abs/2609.13259 · HTML: https://mlsanity.com/paper/2609.13259/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Diffusion and multimodal model alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 26 of its 36 references that are in the map.
- First posted on arXiv 6 Sep 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Diffusion and multimodal model alignment](https://mlsanity.com/alignment/diffusion-and-multimodal-model-alignment/)

## Builds on (most cited first, 26 of 26)

- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency](https://mlsanity.com/paper/2508.18265/) (2025)
- [Flow-GRPO: Training Flow Matching Models via Online RL](https://mlsanity.com/paper/2505.05470/) (2025)
- [VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware Normalization](https://mlsanity.com/paper/2103.16874/) (2021)
- [HPSv3: Towards Wide-Spectrum Human Preference Score](https://mlsanity.com/paper/2508.03789/) (2025)
- [OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-on](https://mlsanity.com/paper/2403.01779/) (2024)
- [TryOnDiffusion: A Tale of Two UNets](https://mlsanity.com/paper/2306.08276/) (2023)
- [Boosting Multimodal Large Language Models with Visual Tokens Withdrawal for Rapid Inference](https://mlsanity.com/paper/2405.05803/) (2024)
- [Improving Diffusion Models for Authentic Virtual Try-on in the Wild](https://mlsanity.com/paper/2403.05139/) (2024)
- [CatVTON: Concatenation Is All You Need for Virtual Try-On with Diffusion Models](https://mlsanity.com/paper/2407.15886/) (2024)
- [Q-Insight: Understanding Image Quality via Visual Reinforcement Learning](https://mlsanity.com/paper/2503.22679/) (2025)
- [RewardDance: Reward Scaling in Visual Generation](https://mlsanity.com/paper/2509.08826/) (2025)
- [EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling](https://mlsanity.com/paper/2509.23909/) (2025)
- [Any2AnyTryon: Leveraging Adaptive Position Embeddings for Versatile Virtual Clothing Tasks](https://mlsanity.com/paper/2501.15891/) (2025)
- [EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing](https://mlsanity.com/paper/2509.26346/) (2025)
- [MLLMs are Deeply Affected by Modality Bias](https://mlsanity.com/paper/2505.18657/) (2025)
- [FireRed-Image-Edit-1.0 Technical Report](https://mlsanity.com/paper/2602.13344/) (2026)
- [SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learning](https://mlsanity.com/paper/2512.24330/) (2025)
- [Learning Flow Fields in Attention for Controllable Person Image Generation](https://mlsanity.com/paper/2412.08486/) (2024)
- [VQ-Insight: Teaching VLMs for AI-Generated Video Quality Understanding via Progressive Visual Reinforcement Learning](https://mlsanity.com/paper/2506.18564/) (2025)
- [PromptDresser: Improving the Quality and Controllability of Virtual Try-On via Generative Textual Prompt and Prompt-aware Mask](https://mlsanity.com/paper/2412.16978/) (2024)
- [BaseReward: A Strong Baseline for Multimodal Reward Model](https://mlsanity.com/paper/2509.16127/) (2025)
- [OmniTry: Virtual Try-On Anything without Masks](https://mlsanity.com/paper/2508.13632/) (2025)
- [FastFit: Accelerating Multi-Reference Virtual Try-On via Cacheable Diffusion Models](https://mlsanity.com/paper/2508.20586/) (2025)
- [Tstars-Tryon 1.0: Robust and Realistic Virtual Try-On for Diverse Fashion Items](https://mlsanity.com/paper/2604.19748/) (2026)
- [Qwen-Image-2.0-RL Technical Report](https://mlsanity.com/paper/2606.27608/) (2026)

## Abstract (arXiv)

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.