Tencent Releases SCoPE for Camera Control in Video Models
multimodalart · x · 2026-08-19
Tencent released SCoPE on Hugging Face, a new method for camera control in video generation models. It allows users to feed a single image and draw a 3D camera path to generate a video shot following that trajectory. A demo is available on Hugging Face Spaces.
Related event: SCoPE Enables Camera-Controlled Video Generation from a Single Image(2 posts)→
More from Multimodal
- MAI-Image-2.6 ranks 3rd on Image Arena; new Voice Chat playground launches — testingcatalog · 2026-08-19
- New short film 'Noise to Signal' created entirely with Invideo's Agent Two — gen_ericai · 2026-08-19
- AI music video created with Pinokio and Maestro tools — cocktailpeanut · 2026-08-19
- Horrifying Scene Demo Stuns, Minimax H3 Video Generation Sparks Debate — cocktailpeanut · 2026-08-19
- Visual library compares 286 Krea 2 styles across 8 reference scenes — matcheal · 2026-08-19
- Seedance 2.5 goes live globally on CapCut with 1080p output for ad-variant workflows — AIwithGhotai · 2026-08-19