Improved Training Methods for Multi-Talker Speech Processing = Treeningmeetodid mitme rääkijaga kõne töötluseks
author
supervisor
statement of authorship
Joonas Kalda ; [supervisor: Tanel Alumäe ; Tallinn University of Technology, School of Information Technologies, Department of Software Science]
type of dissertation
doktoritöö
university/scientific institution
Tallinna Tehnikaülikool
location of publication
Tallinn
publisher
year of publication
pages
153 p. : ill
series
Tallinn University of Technology. Doctoral thesis = Tallinna Tehnikaülikool. Doktoritöö ; 15/2026
subject term
subject of form
ISSN
2585-6898
2585-6901 (PDF)
ISBN
978-9916-80-464-3 (PDF)
978-9916-80-465-0
notes
Autori publikatsioonide nimekiri leheküljel 8
Bibliograafia lehekülgedel 59-70
Kokkuvõte eesti keeles
Kättesaadav võrguteavikuna
Autori CV inglise ja eesti keeles, lk. 150-153
Thesis (Ph.D. in Computer Science) : Tallinn University of Technology, 2026
url
Open Access
Open Access
scientific publication
teaduspublikatsioon
classifier
TalTech department
language
inglise
- Collar-aware training for streaming speaker change detection in broadcast speech
- PixIT: joint training of speaker diarization and speech separation from real-world multi-speaker recordings
- TalTech-IRIT-LIS speaker and language diarization systems for DISPLACE 2024
- Design choices for PixIT-based speaker-attributed ASR: Team ToTaTo at the NOTSOFAR-1 challenge
- Diarization-Guided Multi-Speaker Embeddings
Kalda, J. Improved Training Methods for Multi-Talker Speech Processing = Treeningmeetodid mitme rääkijaga kõne töötluseks. Tallinn : TalTech Press, 2026. 153 p. : ill. (Tallinn University of Technology. Doctoral thesis = Tallinna Tehnikaülikool. Doktoritöö ; 15/2026).