One photo, 30-second K-pop dance: Seedance 2.5 keeps the face locked throughout
aftahi_ai · x · 2026-09-21
- ByteDance's Seedance 2.5 generated a full 30-second K-pop dance performance from a single photo
- The face stays consistent throughout and the choreography remains locked, avoiding the face drift and limb glitches typical of earlier AI dance videos
- The poster calls it possibly the most impressive AI video of 2026, marking a shift from prompt-gambling to reliable output for AI dance content
Related event: Seedance 2.5 Turns a Single Photo into a Full K-Pop Dance Video(5 posts)→
More from Multimodal
- NVIDIA's Déjà View: 117M looped transformer matches models 10x larger at 3D reconstruction — NVIDIAAI · 2026-11-12
- Midjourney V9 rumored to launch soon with images, 3D and vector support — aziz4ai · 2026-11-12
- Adobe embeds generative AI in Premiere and Firefly with a built-in creative agent — emmanuelvivier · 2026-11-12
- Grok can now generate images in-app, and users are already making benchmark chart memes — ChrisUniverse · 2026-11-12
- A LoRA Is All You Need to Turn a Speech LLM Full-Duplex — kastnerkyle · 2026-11-12
- Runway teases triple bet: real-time avatars, interactive worlds, robot action models — c_valenzuelab · 2026-11-12