ASMSP-P2: Automatic Speech Recognition and Speech Emotion Recognition
Tue, 1 Sep, 14:00 - 15:40
Location: Balcony I West
Session Type: Poster
Session Chair: Mathew Magimai Doss, Idiap Research Institute
Track: 1: ASMSP - Acoustic, Speech and Music Signal Processing
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.1: LEVERAGING UNCERTAINTY IN VISUAL SPEECH RECOGNITION: ERROR ESTIMATION AND ACTIVE LEARNING FOR SPEAKER ADAPTATION
Hayata Yuasa, Keitaro Tanaka, Shigeo Morishima, Waseda University, Japan
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.2: PHONEME-FIRST PREDICTION FOR LLM-BASED SPEECH RECOGNITION
Jakob Poncelet, Hugo Van hamme, KU Leuven, Belgium
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.3: Causality Induced Transformer Attention Decoder for Improved ASR
Vivek Tyagi, Oracle AI, Ireland
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.4: Pruning as Regularization: Sensitivity-Aware One-Shot Pruning in ASR
Julian Irigoyen, Danske Bank, Denmark; Arthur Söhler, Copenhagen Business School, Denmark; Andreas Søeborg Kirkedal, Jabra (GN Group), Denmark
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.5: Differential Gated Conformer
Rajul Acharya, Atharva Suryawanshi, Yash Saravane, Cerence AI, India
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.6: Cross-Modal Alignment of Speech Encoders and Indic LLMs for Hindi-English Code-switching ASR
Puneet Singh Bhooi, Vinayak Abrol, IIIT-Delhi, India
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.7: BLIND, MICROSCOPIC METRICS OF HUMAN SPEECH INTELLIGIBILITY USING END-TO-END SPEECH RECOGNITION
Alexander Wißmann, Technische Universität Berlin, Germany; Jasmin Riegel, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany; Steffen Zeiler, Technische Universität Berlin, Germany; Tobias Reichenbach, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany; Dorothea Kolossa, Technische Universtät Berlin, Germany
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.8: TACKLING CROSS-LINGUAL GENERALIZATION IN SPEECH EMOTION RECOGNITION VIA LINGUISTIC GROUPING: A SPOTLIGHT ON NEO-LATIN LANGUAGES
Francesco Dal Ri, Nicola Garau, Nicola Conci, University of Trento, Italy
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.9: COLOR-BASED EMOTION REPRESENTATION FOR SPEECH EMOTION RECOGNITION
Ryotaro Nagase, Ryoichi Takashima, Yoichi Yamashita, Ritsumeikan University, Japan
Tue, 1 Sep, 14:00 - 15:40

ASMSP-P2.10: Multiclass Fairness Analysis of Qwen2-Audio in Speech Emotion Recognition
Federico D'Asaro, Politecnico di Torino, Italy; Juan Jose Marquez Villacis, LINKS Foundation, Italy; Andrea Bottino, Politecnico di Torino, Italy; Giuseppe Rizzo, LINKS Foundation, Italy