Repair images by inpainting masked areas

Analyze images and videos to detect objects

Generate speech from text or files

Find similar images from a collection

The Metropolitan Museum of Art Collection

Art Institute of Chicago Gallery

Generate and manipulate anime images and videos

Enhance images using color-specific tone curves

Open Agent Leaderboard

FitDiT is a high-fidelity virtual try-on model.

Generate text based on input prompts

This is for learning purpose, don't take it seriously :)

Detect and annotate poses in images and videos

Turn videos into 3D portraits from any angle

Display ranked leaderboard for models and RAG systems

Repair images by inpainting missing parts

statistics analysis for linear regression

Versatile audio super resolution (any -> 48kHz) with AudioSR

Analyze images and videos to identify objects

Synthpose Markerless MoCap VitPose