What it does

CatPapers is a hand-curated list of academic papers in computer graphics, computer vision, and machine learning that produce experimental results involving cats. Each entry includes a teaser image, author list, venue, and links to the paper and often the project page. The maintainer accepts additions or removals via email.

The interesting bit

The curation criterion is delightfully narrow: the paper must demonstrate its method on a cat. This turns a sprawling field into a browsable gallery where diffusion models, GANs, video completion, and 3D printing research all share one furry denominator. It is arguably the most honest literature review in vision research.

Key highlights

• Covers work from 1954 (Attneave) through 2025 (diffusion collapse recovery)
• Heavy hitters present: StyleGAN, BANMo, REPA, Neural Best-Buddies
• Each paper shown with thumbnail teaser and full citation
• Companion webpage and CSV export available
• Maintained by Jun-Yan Zhu (CMU)

Caveats

• No search, filtering, or tagging beyond scrolling the chronologically ordered HTML tables
• Submission process is manual email, not a PR workflow

Verdict

Worth a bookmark for vision researchers who want to sanity-check whether a method actually works on something other than ImageNet dogs. Everyone else: a fun five-minute browse, then close the tab.