WE1.AUD.4
SOUND EVENT DETECTION WITH SOFT LABELS: A NEW PERSPECTIVE ON EVALUATION
Manu Harju, Irene Martín-Morató, Toni Heittola, Annamaria Mesaros, Tampere University, Finland
Session:
WE1.AUD: Detection and Classification for Audio and Speech I Lecture
Track:
ASMSP - Acoustic, Speech and Music Signal Processing
Location:
Auditorium
Presentation Time:
Wed, 28 Aug, 11:30 - 11:50 France Time (UTC +1)
Session Chair:
Stefan Goetze, University of Sheffield
Presentation
Discussion
Resources
No resources available.
Session WE1.AUD
WE1.AUD.1: Invariance-based layer regularization for sound event detection
David Perera, Slim Essid, Gaël Richard, Télécom Paris, France
WE1.AUD.2: Correlation of Frechet Audio Distance With Human Perception of Environmental Audio Is Embedding Dependent
Modan Tailleur, LS2N, CNRS, Centrale Nantes, Nantes University, France; Junwon Lee, Gaudio Lab, Inc. / KAIST, Korea (South); Mathieu Lagrange, LS2N, CNRS, Centrale Nantes, Nantes University, France; Keunwoo Choi, Gaudio Lab, Inc., Korea (South); Laurie Heller, Carnegie Mellon University, United States; Keisuke Imoto, Doshisha University, Japan; Yuki Okamoto, Ritsumeikan University, Japan
WE1.AUD.3: Target-specific Dataset Pruning for Compression of Audio Tagging Models
Alexander Werning, Reinhold Häb-Umbach, Paderborn University, Germany
WE1.AUD.4: SOUND EVENT DETECTION WITH SOFT LABELS: A NEW PERSPECTIVE ON EVALUATION
Manu Harju, Irene Martín-Morató, Toni Heittola, Annamaria Mesaros, Tampere University, Finland
WE1.AUD.5: Refining knowledge transfer on audio-image temporal agreement for audio-text cross retrieval
Shunsuke Tsubaki, Doshisha University, Japan; Daisuke Niizumi, Daiki Takeuchi, Yasunori Ohishi, Noboru Harada, NTT Corporation, Japan; Keisuke Imoto, Doshisha University, Japan
WE1.AUD.6: OUTLIER EXPOSURE WITH EFFICIENT DIVISION OF POSITIVE AND NEGATIVE EXAMPLES FOR ANOMALOUS SOUND DETECTION
Yuuki Tachioka, Denso IT Laboratory, Japan