Full Description
The two-volume set LNAI 15299 and 15300 constitutes the refereed proceedings of the 26th International Conference on Speech and Computer, SPECOM 2024, held in Belgrade, Serbia, during November 25-28, 2024.
The 53 full papers included in these proceedings were carefully reviewed and selected from 90 submissions. The book also contains two invited talks in full paper length. The papers are organized in the following topical sections:
Volume I:
Invited papers; automatic speech recognition; speech and language resources; speech synthesis and perception; and speech processing for medicine.
Volume II:
Computational paralinguistics; affective computing; speaker recognition; digital speech processing; natural language processing.
Contents
Invited Papers.- Preserving Language Heritage Through Speech Technology: The Case of Upper Sorbian.- Retrospective and Perspectives of TTS & STT Technology Development and Implementation for South Slavic Under-Resourced Languages.- Automatic Speech Recognition.- Comparison of Well- and Lower-Resourced Self-Training in ASR.- Towards a Livvi-Karelian End-to-End ASR System.- Advances in OpenASR21 Evaluation with Increased Temporal Resolution for Speech Self-Supervised Learning Models.- Benchmarking Whisper under Diverse Audio Transformations and Real-time Constraints.- AutoMode-ASR: Learning to Select ASR Systems for Better Quality and Cost.- Pre-Training and Adverse Audio Samples for Data-Efficient Wake Word Detection.- Cross-Lingual Summarization of Speech-to-Speech Translation: A Baseline.- Speech and Language Resources.- The ParlaSpeech Collection of Automatically Generated Speech and Text Corpora from Parliamentary Proceedings.- ESC Corpus of Spoken Russian: Everyday Student Conversations Captured through Continuous Speech Recording in Natural Communicative Environments.- OpenAV: Bilingual Dataset for Audio-Visual Voice Control of a Computer for Hand Disabled People.- Bulgarian Speech Resources in the CHILDES System.- Multiword Units in Russian Everyday Speech: Empirical Classification and Corpus-Based Studies.- Neurophysiological Correlates of Textual Modulation in Visual Stimuli: An Experimental Study of Russian and English Memes.- Speech Synthesis and Perception.- End-to-End Speech Synthesis for the Serbian Language Based on Tacotron.- ChildTinyTalks (CTT): A Benchmark Dataset and Baseline for Expressive Child Speech Synthesis.- Multidimensional Rhythm: Comparing Rhythmic Properties of Australian and New Zealand Monologues.- Influence of Linguistic and Sociolinguistic Factors on Speech Rate Perception.- Human and Machine Keyphrase Perception in Russian Text and Speech.- Assessment of Children's Ability to Manifest Emotions in Facial Expressions, Voice and Speech by Humans, Automatic, and on a Likert Scale.- Speech Processing for Medicine.- Investigating the Utility of wav2vec 2.0 Hidden Layers for Detecting Multiple Sclerosis.- Cross-Cultural Automatic Depression Detection based on Audio Signals.- Depression Classification using Token Merging-based Speech Spectrotemporal Transformer.- Detecting Depression from Audio Data.- Binary and Multiclass Classification of Dysphonia Using Whisper Encoder and One-Dimensional Convolutional Neural Network.- Approach to Assessing the Quality of Syllable Pronunciation by Patients in the Process of Speech Rehabilitation Based on Comparison with Healthy Speakers.- A Comparative Study for Contextualized Spoken Answer Classification in German Medical Questionnaires.
