An open-source state-of-the-art toolbox for broadcast news diarization

By Mickael Rouvier, G. Dupuy, P. Gay, E. Khoury, Téva Merlin, Sylvain Meignier. 2013

168 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Natural Language Processing, Speech & Audio

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jan Niehues
2. Ngoc-Quan Pham
3. A. Waibel
4. Erica Cooper
5. Sriram Ganapathy
6. Sebastian Stüker
7. Alfons Juan-Císcar
8. Joan Albert Silvestre-Cerdà

In Natural Language Processing

1. Jan Niehues
2. A. Waibel
3. Ngoc-Quan Pham
4. Erica Cooper
5. Sebastian Stüker
6. Alfons Juan-Císcar
7. Joan Albert Silvestre-Cerdà
8. Javier Iranzo-Sánchez

In Speech & Audio

1. Erica Cooper
2. Sriram Ganapathy
3. Sebastian Stüker
4. Benoit Favre
5. Mark Y. Liberman
6. Jordi Luque
7. Jun Du
8. Mickael Rouvier

Citations in the corpus

• VOXLINGUA107: A Dataset for Spoken Language Recognition (2020, methodology): “The LIUM SpkDiarization toolkit [10] was applied for this.”
• Europarl-ST: A Multilingual Corpus for Speech Translation of Parliamentary Debates (2019, methodology): “In order to ameliorate this, first we carried out a Speaker Diarization (SD) step for each speech using the LIUM SpkDiarization [13] toolkit.”
• Text-to-Speech Synthesis Using Found Data for Low-Resource Languages (2019, methodology): “Based on the hypothesis that utterances from similar speakers will create a better voice due to more consistent models, we used the LIUM speaker diarization tool (Rouvier et al., 2013) to cluster our dataset (11094 utterances concatenated into a single audio file of 6 hours and 45 minutes) into grou”
• Speaker diarization through speaker embeddings (2015, methodology): “The diarization system used in experiments is the LIUM Speaker Diarization system [16](2).”
• Audio-to-text alignment for speech recognition with very limited resources (2014, methodology): “For this reason, if the stretched audio is longer than 30 seconds we use the LIUM speaker diarization toolkit [19] to create individual audio files of length less than 30 s, split according to criteria such as silence and, if present, speaker turns.”
• The Second DIHARD Diarization Challenge: Dataset, task, and baselines (2019, background): “, conversational telephone speech [2, 3, 4, 5, 6], broadcast [7, 8], or meeting [9, 10]).”
• Open Source Toolkit for Speech to Text Translation (2018, methodology): “We rely on the LIUM speaker diarization tool (Rouvier et al., 2013) during testing to create the utterances.”