@saf3synt: SV4D 2.0 turns one video into four camera views SV4D 2.0 is a video-to-4D diffusion model that generates 48 frames across four camera views from a single input video. It offers improved motion sharpness and spatio-temporal consistency over the original SV4D. Ideal for researchers working on novel-view synthesis and 4D asset generation. https://github.com/Stability-AI/generative-models #AI #Opensource #StabilityAI #GenerativeVideo Follow for daily open source drops 👇

saf3synt

Open In TikTok:

Region: SI

Sunday 30 August 2026 11:27:55 GMT