Reproduce “Autonomous Fashion Outfit Composition via Unified Aesthetic Foresight Model” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Kaicheng Pang, Xingxing Zou, Ruohan Xu, Waikeung Wong
UAFM composes outfits by retrieving garments and deciding when to stop. Its Transformer pairs a policy head proposing items with a value head, trained by temporal-difference learning, to assess outfits and whether another addition is worthwhile. On curated trajectories, it achieved a 96.2% natural-stop rate versus 34.7% with binary cross-entropy; this includes early stops, while the main benchmark fixed outfit length.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.