PROGRAMME

Jump to: Presentation instructions — Electronic proceedings

The timetable shown below is the final version of the conference programme. Please, check this page regularly for the case there were some changes to the programme based upon particular requests of the conference participants or due to other unforeseen reasons.

Legend: = on-site keynote speech, = online keynote speech, = on-site oral presentation, = online oral presentation, = on-site poster presentation, = online poster presentation

The time of the programme events is given in CEST (Central European Summer Time) or UTC+2.

Monday, 4 September 2023

08:30 – 09:15	Registration (open on all conference days)
09:15 – 09:30	Official Opening of the 26th International Conference TSD2023 — Main Lecture Hall Miloslav Konopík (TSD2023 Organizing Committee Chairman)
09:30 – 10:30	Invited Talk — Main Lecture Hall Philippe Blache: Predictive Coding, Good-enough Processing and Constructions: A Neuro-cognitive Model for Dialogue (link to the recorded presentation)
10:30 – 11:00	Coffee Break
11:00 – 12:40	Oral Session: Speech Synthesis and Quality Evaluation — Main Lecture Hall Chairperson: TBA Daniel Tihelka: VITS: Quality vs. Speed Analysis Frederico Santos de Oliveira: CML-TTS: A Multilingual Dataset for Speech Synthesis in Low-Resource Languages Frederico Santos de Oliveira: Evaluation of Speech Representations for MOS Prediction Juan Camilo Vasquez: When Whisper Meets TTS: Domain Adaptation Using Only Synthetic Speech Data
12:40 – 13:40	Lunch
13:40 – 14:40	Poster Session: Text and Corpus Processing and Analysis — Main Lecture Hall Tomáš Jelínek: Morphological Tagging and Lemmatization of Spoken Corpora of Czech František Trebuňa: Searching for Reasons of Transformers’ Success: Memorization vs. Generalization Hynek Kydlíček: A Dataset and Strong Baselines for Classification of Czech News Texts Noémi Vadász: Resolving Hungarian Anaphora with ChatGPT — Online <log in to see links> Natalia Kalashnikova: The Effect of Human-Likeliness in French Robot-Directed Speech: A Study of Speech Rate and Fluency Kai Hartung: Measuring Sentiment Bias in Machine Translation Győző Zijian Yang: Mono- and Multilingual GPT-3 models for Hungarian
14:40 – 16:20	Oral Session: Text Processing and Summarization — Main Lecture Hall Chairperson: TBA Zdeněk Žabokrtský: The Unbearable Lightness of Morph Classification György Orosz: Advancing Hungarian Text Processing with HuSpaCy: Efficient and Accurate NLP Pipelines Aleš Žagar: One Model to Rule Them All: Ranking Slovene Summarizers Takehito Utsuro: Japanese How-to Tip Machine Reading Comprehension by Multi-task Learning based on Generative Model — Online <log in to see links>
16:20 – 16:50	Coffee Break
17:00 – 22:00	Pilsner Urquell Brewery Excursion & Beer Welcome Party Meeting point: In front of the Pilsner Urquell Brewery Visitors' Centre

Tuesday, 5 September 2023

09:00 – 10:00	Invited Talk — Main Lecture Hall Ivan Habernal: Towards Privacy-Preserving Natural Language Processing (link to the recorded presentation)
10:00 – 10:30	Coffee Break
10:30 – 12:10	Oral Session: Corpora and Language Resources — Main Lecture Hall Chairperson: TBA Gregor Donaj: ParaDiom – A Parallel Corpus of Idiomatic Texts Denis Nikolaus Memmesheimer: A German Parallel Clausal Coordinate Ellipsis Corpus that Aligns Sentences from the TüBa-D/Z Treebank with Reconstructed Canonical Forms Santiago Andres Moreno-Acevedo: Language Generalization using Active Learning in the context of Parkinson’s Disease Classification Juan Manuel Martín-Doñas: An Online Diarization Approach for Streaming Applications Based on Tree-Clustering and Bayesian Resegmentation
12:10 – 13:10	Lunch
13:10 – 14:10	Invited Talk — Main Lecture Hall Juan Rafael Orozco-Arroyave: Speech and Language Markers in Neurodegeneration (link to the recorded presentation)
14:10 – 15:10	Poster Session: Speech Processing and Analysis — Main Lecture Hall José Vicente Egas-López: Identifying Subjects Wearing a Mask from the Speech by Means of Encoded Speech Representations František Kynych: Online Speaker Diarization Using Optimized SE-ResNet Architecture Jan Nouza: Developing State-of-the-Art End-to-End ASR for Norwegian Swayambhu Nath Ray: Unified Modeling of Multi-Domain Multi-Device ASR Systems Jonathan Chevelu: Voice Cloning for Voice Disorders: Impact of Phonetic Content Georgios Karakasidis: Multilingual TTS Accent Impressions for Accented ASR Jan Lehečka: Transfer Learning of Transformer-based Speech Recognition Models from Czech to Slovak Cristian David Rios Urrego: Automatic Pronunciation Assessment of Non-native English based on Phonological Analysis
15:10 – 15:40	Coffee Break
15:40 – 17:20	Oral Session: Speech Recognition and Summarization — Main Lecture Hall Chairperson: TBA Raul Pombo Monteiro: Towards End-to-end Speech-to-text Summarization Lars Formoe: Unsupervised Learning for Automatic Speech Recognition In Air Traffic Control Environment Thibault Bañeras-Roux: HATS: An Open data set Integrating Human Perception Applied to the Evaluation of Automatic Speech Recognition Metrics Tobias Weise: Impact of Including Pathological Speech in Pre-Training on Pathology Detection	Programme Committee Meeting — Lounge & Online (TSD Programme Committee members only)
18:00 – ...	Closing Ceremony of the 26th International Conference TSD2023 — Main Lecture Hall Kamil Ekštein, Miloslav Konopík, Roman Mouček (TSD2023 Organizing Committee Members)

Wednesday, 6 September 2023

09:45 – 16:00

Whole-day Conference Trip to the West-Bohemian Baroque Countryside
Meeting point: In front of the Primavera Hotel

PRESENTATION INSTRUCTIONS

ORAL PRESENTATION

The time for presentation is 20 minutes. Afterwards, a 5 minutes time span is left for a discussion.
Please, prepare your presentation in either Adobe PDF format or Microsoft PowerPoint format if you plan to use the computer installed on site (it is an Intel-based PC with Microsoft Windows).
You can use your own computer for the presentation as well. In such case, of course, there is no restriction on the format of the presentation. However, if your computer has a less common display port (like e.g. the DisplayPort), please, bring a converter with you (VGA D-sub 15 and HDMI are considered standard and work fine with our projectors).
The recommended aspect ratio of the presentation is 16:9 or 16:10.

POSTER PRESENTATION

The posters will be displayed in a designated conference hall/hallway using stand-alone pinboards. They have neutral grey frames and dark blue fabric backgrounds. Pins and a duct tape (transparent) will be available on spot from the organizers.
Each board is portrait-oriented and the dimensions of its utilizable area are: width 865 millimetres (approx. 34 in), height 1175 millimetres (approx. 46 in).
The recommended ISO 216/DIN 476 format of posters is A1, portrait-oriented. A0 paper must be cropped (by 14 mm of height). The largest ISO 216/DIN 476 formats that fit onto the boards without cropping are thus A1 and C1. The largest U.S./Canadian/imperial formats are ANSI/ASME Y14.1 size D, Arch D, and Atlas or Double Demy.
Your poster needs to be installed at least 20 minutes before the beginning of the poster session (the designated hall will be accessible for the poster authors).

NEW PROCEEDINGS

The online book of TSD2023 proceedings is only available to the paying participants of the conference.
If you are one, please, log in in order to access the material!