LTX-2.5 BBox Control enables precise video generation via bounding boxes
linoy_tsaban · x · 2026-08-26
LTX-2.5 introduces BBox Control, allowing users to precisely position subjects in generated videos by drawing bounding boxes and describing content. The feature is available as a demo on Hugging Face Spaces.
More from Multimodal
- Pika releases WAN 3.0 Prime: 7x faster generation with same quality — aziz4ai · 2026-08-27
- With modern prompt adherence, you can deliberately recreate Deep Dream-era rough aesthetics — Merzmensch · 2026-08-27
- Viso Now simplifies computer vision to LLM-style prompting — heyronir · 2026-08-27
- Manhattan at 4 AM: Video Generated with Minimax H3 Tested — cocktailpeanut · 2026-08-27
- Open-source Gepard TTS beats 24 closed APIs with 68ms latency on RTX 4090 — ylankgz · 2026-08-27
- LAION releases 10M-hour open video dataset BVD for multimodal training — HirokatuKataoka · 2026-08-27