MiDashengLM-Gen: Unified Audio Scene Generation via LLM
fruesome · reddit · 2026-08-16
MiDashengLM-Gen is an end-to-end framework using a pre-trained LLM and audio tokenizer. It employs per-token conditional flow matching for autoregressive, variable-length audio scene generation, blending speech, music, SFX, and acoustics coherently from text descriptions.
More from Multimodal
- How meloform works: WASM pitch detection plus differentiable circuits learning 1,022 sine coefficients — henloitsjoyce · 2026-10-16
- Documentary filmmaker urges peers to watch 'You Can See Everything' — kyliebytes · 2026-10-16
- Redditor's first Cumfy/Seedance video nails a cinematic one-shot generation — Wave_N_Wind · 2026-10-16
- Kling 4 returns after 8 months: 10 hyper-realistic AI ads for under $5 — CurieuxExplorer · 2026-10-16
- Shanghai AI Lab unveils Intern Lumina U2, a diffusion LLM unifying image, video and 3D tasks — Shanghai-AI-Laboratory · 2026-10-16
- NAVER AI Lab proposes Speculative Forcing to speed up autoregressive video generation — naver-ai · 2026-10-16