CamDistill Camera Motion Demo
Temporally grounded camera-motion analysis for video clips
None defined yet.
Temporally grounded camera-motion analysis for video clips
Text-to-video world model (EVOKE-Turbo, 14B DiT, 3-step)
Typed questions in, calibrated probabilities out
Official prompt rewriters for Qwen-Image-2.1, kept warm
Generate and edit images with Qwen-Image-2.1
Thai/English decision model - calibrated answers, no text
Closed-loop LIBERO rollouts from a world-action policy
Precise regional image editing with MaskFlow
A fruit fly connectome writing bedtime stories
Design sheets to game cutscenes with HUD and audio
4-step MeanFlow Japanese TTS with voice cloning
Typed questions in, calibrated probabilities out
Live 3D demo of Google's GNM parametric head model
Answer typed questions about a text in one forward pass
Speak to Realtime-Venus, get text + speech back
Typed decisions with calibrated probabilities, one pass
8-step MiniMax H3 video+audio with the HyperFlow LoRA
1970s hand-painted cel animation, with its own soundtrack
Anime art from structured Danbooru captions
One-pass calibrated decisions from an image
Typed decisions with probabilities from text + JSON schema
SNES-era pixel-art video with sound, from text
Industrial rock songs from lyrics with a YuE2 LoRA
Turn document page images into structured Markdown