4Director Lets You Direct AI Video by Placing Cameras and Objects in 3D
anand_bhattad · x · 2026-10-03
Researchers from JHU (collaborating with Alan Yuille) introduce 4Director: users place the camera and subjects in 3D, draw their paths, and the model generates the shot accordingly, turning directing work like camera moves, staging, and timing into controllable inputs.
The authors note all training and experiments ran on academic compute (supported by Johns Hopkins DSAI); more results are on the project page, and related work appeared the same day.
Related event: 4Director: Rigid 3D Geometry Controls Video World Models(2 posts)→
More from Multimodal
- Developer builds a fully code-generated video using Opus 5.5 and generative image models — leebrimelow · 2026-10-03
- Creator shares a favorite StyleGAN model: "such a trip" — makeitrad1 · 2026-10-03
- HeyGen open-sources HyperFrames: deterministic HTML-to-MP4 video pipelines with an agent skill pack — lmoroney · 2026-10-03
- fal launches MiniMax H3 Max Recast: swap video subjects from a reference photo at $0.30/s, motion and audio intact — lmoroney · 2026-10-03
- Qwen-Image-2.1-viggle-turbo: fast 6-step character sheets in ComfyUI — RiverSide71h · 2026-10-03
- Camera Control for AI Images in ComfyUI with QI 2.1 — RobbaW · 2026-10-03