Karpov / Gosztolya Speech and Computer

27th International Conference, SPECOM 2025, Szeged, Hungary, October 13–15, 2025, Proceedings, Part II
Erscheinungsjahr 2025
ISBN: 978-3-032-07959-6
Verlag: Springer International Publishing
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 347 Seiten

Reihe: Computer Science (R0)

This two-set volume LNAI 16187 and 16188 constitutes the refereed proceedings of the 27th International Conference on Speech and Computer SPECOM 2025 held in Szeged, Hungary, during October 13–15, 2025.

The 47 full papers and 1 invited paper included in this book were carefully reviewed and selected from 77 submissions. The papers are organized in the following topical sections:  

Part I: Invited Paper; Speech Perception and Synthesis; Computational Paralinguistics; Speech Processing for Healthcare; Speech and Language Resources; Speaker Recognition.

Part II:  Automatic Speech Recognition; Speech Processing for Under-Resourced Languages; Digital Speech Processing; Natural Language Processing; Multimodal Systems.

Zielgruppe

Research

Autoren/Hrsg.

Karpov, Alexey

Gosztolya, Gábor

Inhaltsverzeichnis

.- Automatic Speech Recognition.
.- In-Domain SSL Pre-Training and Streaming ASR: Application to Air Traffic Control Communications.
.- Evaluating the Performance of Several ASR Systems in Environmental and Industrial Noise.
.- Ground Truth-Free WER Prediction for ASR via Audio Quality and Model Confidence Features.
.- Enhancing Speech Recognition through Text-to-Speech and Voice Conversion Augmentation.
.- Best Data is more Supervised Data - Even for Hungarian ASR.
.- Arabic ASR on the SADA Large-Scale Arabic Speech Corpus with Transformer-based Models.
.- Speech Processing for Under-Resourced Languages.
.- Effect of Increased Temporal Resolution on Speech Recognition for French Quebec using Features from Speech Self-Supervised Learning Models.
.- Modeling Intra-Word Code-Switching for Karelian ASR.
.- Improving Whisper-based Serbian ASR using Synthetic Speech.
.- Domain Knowledge and Language Embeddings for Low-Resource Multilingual Phoneme ASR.
.- Whistler Identification in Whistled Spanish (Silbo): A Case Study.
.- Digital Speech Processing.
.- PinkVocalTransformer: Neural Acoustic-to-Articulatory Inversion based on the Pink Trombone.
.- CrossMP-SENet: Transformer-based Cross-Attention for Joint Magnitude-Phase Speech Enhancement. 
.- Adaptive Singing Voice Enhancement for Live Stages.
.- Revealing the Hidden Temporal Structure of HubertSoft Embeddings based on the Russian Phonetic Corpus.
.- Natural Language Processing.
.- Analyzing Web-Scraped and Generated Inputs for Automatic and Scalable Intent Classification.
.- Enhancing Retrieval Performance via LLM Hard-Negative Filtering.
.- Sector-Wise Backpropagation for Low-Resource Text Classification in Deep Models.
.- High-Frequency Multiword Units and the Typological Distribution of Multiword Units in Spoken Russian.
.- Estimation of the Genre Composition of the English Subcorpus of the Google Books Ngram.
.- Multimodal Systems.
.- Ensembling Synchronisation-based and Face-Voice Association Paradigms for Robust Active Speaker Detection in Egocentric Recordings.
.- Phonetic and Visual Characteristics of Cognitive Load.
.- Cognitive Humor Processing in the Russian and English Internet Meme Chatting: EEG Study.
.- Saudi Sign Language Translation Using T5.