Hakim Missoum
Publications
Woosh: A Sound Effects Foundation Model
ARXIV, 2026 | Gaëtan Hadjeres, Marc Ferras, Khaled Koutini, Benno Weck, Alexandre Bittar, Thomas Hummel, Zineb Lahrici, Hakim Missoum, Joan Serrà, Yuki Mitsufuji
Woosh is Sony AI's open sound effects foundation model featuring high-quality audio encoding, text-to-audio, and video-to-audio generation. Optimized for sound effects, it offers competitive performance against models like StableAudio-Open and TangoFlux, with distilled models for fast, low-resource inference.