← all repositories
lzhbrian/Cool-Fashion-Papers

A curated closet of fashion-AI research

Someone finally organized the chaos of virtual try-on papers, outfit recommenders, and trend forecasters into one browsable list.

Cool-Fashion-Papers
Velocity · 7d
+0.2
★ / day
Trend
steady
star history

What it does This is a hand-curated awesome-list tracking AI research at the intersection of computer vision and fashion. It catalogs papers, datasets, workshops, and companies across four main tracks: image synthesis (virtual try-on, garment transfer), classification (landmark detection, brand recognition), recommendation (outfit compatibility, personalized generation), and trend forecasting.

The interesting bit The maintainer sorts papers by arXiv first-submission date—a small but useful discipline that keeps the chronology honest and makes it easy to trace how virtual try-on evolved from coarse warping (VITON, 2018) to high-resolution detail preservation (VITON-HD, Pose with Style, 2021). The list also surfaces industry-facing workshops and challenges that academic search engines often bury.

Key highlights

  • ~60+ papers with direct arXiv or conference PDF links, plus code repositories where available
  • Covers CVPR, ICCV, ECCV, SIGGRAPH, KDD, RecSys, and domain-specific workshops
  • Datasets section includes Fashionpedia, DeepFashion2, and FashionGen
  • Related events list spans 2016–2020, including KDD Fashion workshops and CVPR creative-vision tracks
  • Companies section (logos in the repo) names firms like Stitch Fix, Zalando, Heuritech, and Syte

Caveats

  • README is a flat table of contents with no search, tagging, or filtering—purely manual browsing
  • “Constantly updating” per the description, but last visible workshop entries stop around 2020; freshness beyond that is unclear
  • No annotation or quality ranking beyond inclusion

Verdict Worth bookmarking if you’re entering fashion-AI research and need a quick map of the territory. Skip it if you want interactive exploration or systematic literature review tools—this is a bibliography with URLs, not a database.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.