RelateAnything
Open-vocabulary scene graphs from any image regions
None defined yet.
Open-vocabulary scene graphs from any image regions
Play chess against a 230M-parameter language model
Video continuation with an object-permanence world model
Text-promptable medical image segmentation with SAM 3
Train a tabular foundation model in one forward pass
Pronoun-aware English to German translation with context
Chat with an 11B omni model that replies as a talking avatar
Articulate an object from one rest-state image
8-camera multi-task driving perception, 12 heads at once
Steer an infinite interactive world rollout live
Diffusion model that writes renderable UI screens
0.2M-param 3-class sentiment classifier
8-step MiniMax-H3 β video with a matching soundtrack
AI cover songs from reference audio, lyrics and style
6-DoF pose of novel objects from RGB-D plus a CAD model
4-step video + native stereo audio from LynnReal-Omni
Instrumental music from style tags + a section plan
Controllable music generation with audio conditioning
Ophthalmic multimodal reasoning on retinal images
Spatial & visual reasoning VLM β images and video
LTX-2.5 video with audio from text or an image
Unified text-to-image generation and image editing demo
Moved β use the authors' Space dipta007/OracleZoom
Assess spatial aesthetics of interior images