MetaEncoder
Multimodal decision making and retrieval with one encoder
None defined yet.
MIMESIS: Learning User Simulators as Training Environments for Interactive Agents
Scaling Laws for Looped Mixture of Experts
Multimodal decision making and retrieval with one encoder
Submit and view model evaluation on molecular benchmarks
Generate music from text and optional melody
Generate short music clips from text prompts
Generate animated 3D meshes from video
Track points in a video
Run molecular dynamics or relaxation simulations with UMA
Submit and validate Wearable‑AI challenge predictions
Massively multilingual translation benchmark
Explore multilingual agent benchmark results
Explore and filter data with an interactive dashboard
Reconstruct 3D scenes from photos or videos
On-Device Track Anything Model
Show the SA‑3DAO leaderboard rankings
Generate artist-like meshes from point clouds with MeshFlow
3D reconstruction from images/video with VGGT-Omega
Extract soft human matte and foreground from an image
Generate detailed 308‑point human pose visualizations from images
Segment human body parts in an image into 29 classes
Generate per-pixel surface normal maps from photos
Generate a 3D point‑cloud mesh from a single image
VGGT (CVPR 2025)
Dense Point Tracking with Cost-Volume Free Warping
Leaderboard for watermarking models