SS-L8: Promise and Perils of Generative AI for the Evaluation of Speech and Audio
Oral
Wed, 6 May, 16:30 - 18:30
Location: Room 116
Session Type: Oral
Track: Special Sessions
Click the to view the manuscript on IEEE Xplore Open Preview
Wed, 6 May, 16:30 - 16:50

SS-L8.1: Assessing speech quality metrics for evaluation of neural audio codecs under clean speech conditions
Wolfgang Mack, Cisco Systems Inc., Germany; Nezih Topaloglu, Yeditepe University, Türkiye; Laura Lechler, Ivana Balić, Alexandra Craciun, Mansur Yesilbursa, Kamil Wojcicki, Cisco Systems Inc., Türkiye
Wed, 6 May, 16:50 - 17:10

SS-L8.2: DISTILMOS: LAYER-WISE SELF-DISTILLATION FOR SELF-SUPERVISED LEARNING MODEL-BASED MOS PREDICTION
Jianing Yang, Wataru Nakata, Yuki Saito, Hiroshi Saruwatari, The University of Tokyo, Japan
Wed, 6 May, 17:10 - 17:30

SS-L8.3: Evaluating ASR–LLM Setups for Japanese Speech Recognition with Multi-Pass Augmented Generative Error Correction
Yuka Ko, Karlsruhe Institute of Technology (KIT), Germany; Sheng Li, Institute of Science Tokyo, Japan; Chao-Han Huck Yang, NVIDIA Research, Taiwan; Tatsuya Kawahara, Kyoto Univerisity, Japan
Wed, 6 May, 17:30 - 17:50

SS-L8.4: ARE THESE EVEN WORDS? QUANTIFYING THE GIBBERISHNESS OF GENERATIVE SPEECH MODELS
Danilo de Oliveira, Tal Peer, Jonas Rochdi, Timo Gerkmann, University of Hamburg, Germany
Wed, 6 May, 17:50 - 18:10

SS-L8.5: 2025 URGENT SPEECH ENHANCEMENT CHALLENGE MULTILINGUAL P.808 LISTENING TESTS: APPROACH AND RESULTS
Marvin Sach, Yihui Fu, Technische Universität Braunschweig, Germany; Kohei Saijo, Waseda University, Japan; Wangyou Zhang, Shanghai Jiao Tong University, China; Samuele Cornell, Carnegie Mellon University, United States of America; Robin Scheibler, Google DeepMind, Japan; Chenda Li, Shanghai Jiao Tong University, China; Zhaoheng Ni, Anurag Kumar, Meta, United States of America; Wei Wang, Yanmin Qian, Shanghai Jiao Tong University, China; Shinji Watanabe, Carnegie Mellon University, United States of America; Tim Fingscheidt, Technische Universität Braunschweig, Germany
Wed, 6 May, 18:10 - 18:30

SS-L8.6: FEW-SHOT AND PSEUDO-LABEL GUIDED SPEECH QUALITY EVALUATION WITH LARGE LANGUAGE MODELS
Ryandhimas E. Zezario, Dyah A.M.G. Wisnu, Academia Sinica, Taiwan; Szu-Wei Fu, NVIDIA, Taiwan; Sabato Marco Siniscalchi, University of Palermo, Italy; Hsin-Min Wang, Yu Tsao, Academia Sinica, Taiwan