vivo and Xidian Propose StereoBind for Spatially Consistent Stereo Audio in Video Generation
Hanmo Chen · hf · 2026-10-12
vivo BlueImage Lab and Xidian University propose StereoBind, tackling the overlooked stereo-audio problem in joint video-audio generation via 'Dynamic Spatial Correspondence' — perceived sound location evolving consistently with its visual source's motion.
The framework uses motion tracks through three mechanisms: Visual Motion Binding (source-aware audiovisual correspondence), a Spatial Track Encoder (absolute source positions), and Residual Track RoPE (relative motion).
They also release StereoWorld-29K, a large-scale stereo audio-video dataset with paired motion tracks, plus the StereoWorldBench benchmark. Experiments show substantially improved spatial alignment while preserving overall audiovisual quality.
More from Multimodal
- 'Jamlok', billed as world's first AI-powered horror film, targets 3-month completion — lmoroney · 2026-10-12
- "World's First AI-Powered Horror Film" Jamlok Unveiled in New Delhi, Due in ~3 Months — lmoroney · 2026-10-12
- Minimax H3 faces go blurry in wide shots: ComfyUI users seek fixes — Unique_Bag1444 · 2026-10-12
- Fixing silhouette and palette, exploring only costume: scoping a local SDXL character workflow — Alchemistwiza · 2026-10-12
- LoRA Pilot ep.3: character dataset prep with browser upload, image review and captions — no3us · 2026-10-12
- MiniMax voice cloning trouble: cloned character babbles and speaks too early — Nokersoek · 2026-10-12