5h ago

Just Dub It performs one-pass AI video dubbing

52811114.9K

——0——

Naomi Ken Korem developed Just Dub It, an AI system that performs video dubbing in one pass through joint audio and video processing on the LTX model. The approach handles extreme head profiles and occlusions while generating synchronized output. The work has been accepted to SIGGRAPH 2026. A demonstration shows sequential frames of a child seated on a couch holding pizza. Andrew Carr and Kory Mathewson also posted a linked video and code reference.

Original post

KM#618@KORYMATH @NOAMIKENKOREM

Naomi Ken Korem@NOAMIKENKOREM

1/2 Excited to share my work with @antonch44379293 & Gal Zeevi: Just Dub It! 🎬🔊 We solve dubbing in one pass via joint audio/video on @ltx_model, handling extreme profiles & occlusions with ease. Accepted to SIGGRAPH 2026! 🎓 🧵👇

9:52 AM · May 11, 2026

Cluster engagement

19 snapshots

AI 1000 · 2 actions

QUOTEAC#564@ANDREW_N_CARR @NOAMIKENKOREMcathedrals everywhere https://x.com/andrew_n_carr/status/2053919590993060147/photo/1
REPOSTKM#618@KORYMATH @NOAMIKENKOREM1/2 Excited to share my work with @antonch44379293 & Gal Zeevi: Just Dub It! 🎬🔊 We solve dubbing in one pass via joint audio/video on @ltx_model, handling extreme profiles & occlusions with ease. Accepted to SIGGRAPH 2026! 🎓 🧵👇 https://x.com/noamiKenKorem/status/2053881029723636014/video/1