Woosh: Sony AI Open-Sources Sound Effects Foundation Model

arXiv / Sony AI·low signal

Sony AI released Woosh, an open-source foundation model for sound effects covering audio encoding/decoding, text-audio alignment, text-to-audio, and video-to-audio generation. Includes distilled variants for fast inference and outperforms StableAudio-Open and TangoFlux on benchmarks. Fills an underserved modality gap in open-source generative AI.