ASMSP-SS2: Sound of Change – Advances in Acoustic Anomaly Detection
Thu, 3 Sep, 10:30 - 12:10
Location: Forum 6
Session Type: Lecture
Session Co-Chairs: Tomoya Nishida, Hitachi and Keisuke Imoto, Kyoto University
Track: 20: Sound of Change – Advances in Acoustic Anomaly Detection
Thu, 3 Sep, 10:30 - 10:50

ASMSP-SS2.1: HOW MUCH DOES MACHINE IDENTITY MATTER IN ANOMALOUS SOUND DETECTION AT TEST TIME?
Kevin Wilkinghoff, Aalborg University/Pioneer Centre for AI, Denmark; Keisuke Imoto, Kyoto University, Japan; Zheng-Hua Tan, Aalborg University/Pioneer Centre for AI, Denmark
Thu, 3 Sep, 10:50 - 11:10

ASMSP-SS2.2: MEMNMF: MEMORY-AUGMENTED NMF ON LPC SPECTRA FOR ANOMALOUS SOUND DETECTION
Phurich Saengthong, Takahiro Shinozaki, Institute of Science Tokyo, Japan
Thu, 3 Sep, 11:10 - 11:30

ASMSP-SS2.3: SPIKE BASED APPROACH FOR ANOMALY LOCALIZATION IN LENGTHY BIOACOUSTIC RECORDINGS
Sarangan Raveendranathan, Prasanga N. Samarasinghe, Thushara D. Abhayapala, The Australian National University, Australia
Thu, 3 Sep, 11:30 - 11:50

ASMSP-SS2.4: Generalized Zero-Shot Learning of Acoustic Event Using Attribute-Conditioned GAN and Out-of-Distribution Detector
Kohei Uehara, Kobe University, Japan; Ryoichi Takashima, Ritsumeikan University, Japan; Tetsuya Takiguchi, Kobe University, Japan
Thu, 3 Sep, 11:50 - 12:10

ASMSP-SS2.5: Inter-Audio Diverse Decoding for Audio Difference Captioning in Unsupervised Anomalous Sound Detection Scenarios
Tomoya Nishida, Yusuke Kanamori, Harsh Purohit, Kota Dohi, Takashi Endo, Yohei Kawaguchi, Hitachi, ltd., Japan