ASMSP-L4: Learning-Based Audio Modeling
Thu, 3 Sep, 14:00 - 15:40
Location: Concert Hall
Session Type: Lecture
Session Chair: Annamaria Mesaros, Tampere University
Track: 1: ASMSP - Acoustic, Speech and Music Signal Processing
Thu, 3 Sep, 14:00 - 14:20

ASMSP-L4.1: MUKA: MULTI KERNEL AUDIO ADAPTATION OF AUDIO-LANGUAGE MODELS
Reda Bensaid, IMT Atlantique / Polytechnique Montréal, France; Amine Ouasfi, INRIA, university rennes, France; Yassir Bendou, IMT Atlantique, France; Ilyass Moummad, INRIA, Université de Montpellier, France; Vincent Gripon, IMT Atlantique, France; François Leduc-Primeau, Polytechnique Montréal, Canada; Adnane Boukhayma, INRIA, university rennes, France
Thu, 3 Sep, 14:20 - 14:40

ASMSP-L4.2: What Do Neurons Listen To? A Neuron-level Dissection of a General-purpose Audio Model
Takao Kawamura, Daisuke Niizumi, Nobutaka Ono, Tokyo Metropolitan University, Japan
Thu, 3 Sep, 14:40 - 15:00

ASMSP-L4.3: DISSE: LEARNING DISENTANGLED SOURCE AND SPATIAL REPRESENTATIONS FROM SPATIAL AUDIO–TEXT CONTRASTIVE LEARNING
Shotaro Ueji, Keio University, Japan; Shinnosuke Takamichi, Keio University / The University of Tokyo, Japan; Kouei Yamaoka, The University of Tokyo, Japan
Thu, 3 Sep, 15:00 - 15:20

ASMSP-L4.4: PHYSICS-INFORMED NEURAL ENGINE SOUND MODELING WITH DIFFERENTIABLE PULSE-TRAIN SYNTHESIS
Robin Doerfler, Impulse Audio Lab GmbH, Germany; Lonce Wyse, Universitat Pompeu Fabra, Spain
Thu, 3 Sep, 15:20 - 15:40

ASMSP-L4.5: SylVAD: Improving In-Vehicle Voice Activity Detection via Syllable-Level Alignment
Eunmi Seo, Hyundai Motor Company, Korea (South)