Stable-Layers: Fine-Tuning Image Layer Decomposition Models with VLM-Scored Reinforcement Learning

Ciara Rowles; Reshinth Adithyan; Nikhil Pinnaparaju; Vikram Voleti; Mark Boss

title	Stable-Layers: Fine-Tuning Image Layer Decomposition Models with VLM-Scored Reinforcement Learning
authors	Ciara Rowles, Reshinth Adithyan, Nikhil Pinnaparaju, Vikram Voleti, Mark Boss
date	2026-05-28
venue	arXiv (preprint, not peer reviewed)
id	arXiv:2605.30257
affiliation	Stability AI
upvotes	9 (Hugging Face Daily Papers, 2026-11-01)
links	abs pdf hf
abstract	We present Stable-Layers, a reinforcement learning framework that eliminates the need for paired supervision by fine-tuning a pretrained layer decomposition model using only feedback from a vision-language model (VLM). Starting from Qwen-Image-Layered, we apply Flow-GRPO with LoRA adaptation, sampling multiple candidate decompositions per image, scoring them with a VLM, and optimising the policy from group-relative advantages. The key challenge lies in designing a reliable reward signal: VLMs scoring samples in isolation tend to compress their judgements into a narrow band, leaving GRPO with little within-group variance to learn from. We address this with a two-stage evaluation pipeline that pairs structured per-sample scoring across five edit-centric criteria with a grid-based calibration step in which the VLM re-scores all candidates side-by-side. Stable-Layers produces decompositions with stronger layer separation, fewer blank or artifact-heavy layers, and lower per-layer reconstruction error on the Crello dataset compared to the base model.

cite

@misc{rowles2026stable,
  title={Stable-Layers: Fine-Tuning Image Layer Decomposition Models with VLM-Scored Reinforcement Learning},
  author={Ciara Rowles and Reshinth Adithyan and Nikhil Pinnaparaju and Vikram Voleti and Mark Boss},
  year={2026},
  eprint={2605.30257},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2605.30257}
}