Publications
Quantifying the Role of Textual Predictability in Automatic Speech Recognition
[2024] Proceedings of the Twenthy-Fifth Annual Conference of the International Speech Communication Association (INTERSPEECH 2024).
Sean Robertson, Gerald Penn, Ewan Dunbar
[2024] ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
Kuan-Po Huang, Chih-Kai Yang, Yu-Kuan Fu, Ewan Dunbar, Hung-yi Lee
A morphological analyzer for Huasteca Nahuatl
[2023] Proceedings of the Sixth Workshop on the Use of Computational Methods in the Study of Endangered Languages.
Ana Tona, Guillaume Thomas, Ewan Dunbar
Bigger is not Always Better: The Effect of Context Size on Speech Pre-Training
[2023] arXiv preprint arXiv:2312.01515.
Sean Robertson, Ewan Dunbar
Evaluating context-invariance in unsupervised speech representations
[2023] Proceedings of the Twenty-Fourth Annual Conference of the International Speech Communication Association (INTERSPEECH 2023).
Mark Hallap, Emmanuel Dupoux, Ewan Dunbar
[2023] Society for Computation in Linguistics.
Zhanao Fu, Ewan Dunbar
Are word boundaries useful for unsupervised language learning?
[2022] arXiv preprint arXiv:2210.02956.
Tu Anh Nguyen, Maureen De Seyssel, Robin Algayres, Patricia Roze, Ewan Dunbar, Emmanuel Dupoux
Do self-supervised speech models develop human-like perception biases?
[doi]
[2022] Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).
Juliette Millet, Ewan Dunbar
Predicting non-native speech perception using the Perceptual Assimilation Model and state-of-the-art acoustic models
[doi]
[2022] Proceedings of the 25th Conference on Computational Natural Language Learning.
Juliette Millet, Ioana Chitoran, Ewan Dunbar
Self-supervised language learning from raw audio: Lessons from the zero resource speech challenge
[doi]
[2022] IEEE Journal of Selected Topics in Signal Processing.
Ewan Dunbar, Nicolas Hamilakis, Emmanuel Dupoux
Toward a realistic model of speech processing in the brain with self-supervised learning
[2022] Advances in Neural Information Processing Systems.
Juliette Millet, Charlotte Caucheteux, Pierre Orhan, Yves Boubenec, Alexandre Gramfort, Ewan Dunbar, Christophe Pallier, Jean-Remi King
Paraphrases do not explain word analogies
A reply to Allen and Hospedales (2019, ICML), in which we show that their explanation of why distributional word vectors encode linguistic regularities is empirically incorrect.
[2021] Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics.
Louis Fournier, Ewan Dunbar
The Zero Resource Speech Challenge 2021: Spoken language modelling
Results of the Interspeech edition of the Zero Resource Speech Challenge 2021, a machine learning challenge aimed at learning language models directly from speech
[2021] INTERSPEECH 2021: Twenty-second Annual Conference of the International Speech Communication Association.
Ewan Dunbar, Mathieu Bernard, Nicolas Hamilakis, Tu Anh Nguyen, Maureen Seyssel, Patricia Rozé, Morgane Rivière, Eugene Kharitonov, Emmanuel Dupoux
[zero-resource]
Analogies minus analogy test: measuring regularities in word embeddings
[doi]
[code]
[2020] Proceedings of the 24th Conference on Computational Natural Language Learning (CoNLL).
Louis Fournier, Emmanuel Dupoux, Ewan Dunbar
[representations]
Independent and Automatic Evaluation of Speaker-Independent Acoustic-to-Articulatory Reconstruction
A new method for evaluating how well articulatory reconstruction does at capturing speaker-independent information, using a phone discrimination task.
[2020] INTERSPEECH 2020: Twenty-First Annual Conference of the International Speech Communication Association.
Maud Parrot, Juliette Millet, Ewan Dunbar
Modelling perceptual effects of phonology with ASR systems
[2020] Proceedings for the 42nd Annual Meeting of the Cognitive Science Society.
Bing Er Jiang, Ewan Dunbar, Morgan Sonderegger, Meghan Clayards, Emmanuel Dupoux
Perceptimatic: A human speech perception benchmark for unsupervised subword modelling
[code]
[data]
An open experimental English and French phone discrimination data set which we test on the models submitted to the Zero Resource Speech Challenge 2017, showing that typical automatic speech recognition models show very poor correlation with human speech perception, while many unsupervised models perform much better.
[2020] INTERSPEECH 2020: Twenty-First Annual Conference of the International Speech Communication Association.
Juliette Millet, Ewan Dunbar
[perception]
The Perceptimatic English benchmark for speech perception models
[code]
[data]
[2020] Proceedings for the 42nd Annual Meeting of the Cognitive Science Society.
Juliette Millet, Ewan Dunbar
[perception]
[2020] NeurIPS Workshop on Self-Supervised Learning for Speech and Audio Processing.
Tu Anh Nguyen, Maureen Seyssel, Patricia Rozé, Morgane Rivière, Evgeny Kharitonov, Alexei Baevski, Ewan Dunbar, Emmanuel Dupoux
[zero-resource]
The Zero Resource Speech Challenge 2020: Discovering discrete subword and word units
[doi]
[2020] INTERSPEECH 2020: Twenty-First Annual Conference of the International Speech Communication Association.
Ewan Dunbar, Julien Karadayi, Mathieu Bernard, Xuan-Nga Cao, Robin Algayres, Lucas Ondel, Laurent Besacier, Sakriani Sakti, Emmanuel Dupoux
[zero-resource]
Animal linguistics and the puzzle of titi monkeys alarm sequences
[2019] Proceedings of the 22nd Amsterdam Colloquium, Amsterdam, Netherlands.
Mélissa Berthet, Juan Benjumea, Juliette Millet, Cristiane Cäsar, Klaus Zuberbühler, Ewan Dunbar
Comparing unsupervised speech learning directly to human performance in speech perception
[code]
[data]
[2019] Proceedings for the 41st Annual Meeting of the Cognitive Science Society.
Juliette Millet, Nika Jurov, Ewan Dunbar
[perception]
Generative grammar, neural networks, and the implementational mapping problem: Response to Pater
[doi]
[2019] Language.
Ewan Dunbar
[representations]
Mouse tracking as a window into decision making
[doi]
[2019] Behavior research methods.
Mora Maldonado, Ewan Dunbar, Emmanuel Chemla
RNNs implicitly implement tensor product representations
[2019] International Conference on Learning Representations (ICLR).
R Thomas McCoy, Tal Linzen, Ewan Dunbar, Paul Smolensky
[representations]
The Zero Resource Speech Challenge 2019: TTS without T
[doi]
[2019] INTERSPEECH 2019: Twentieth Annual Conference of the International Speech Communication Association.
Ewan Dunbar, Robin Algayres, Julien Karadayi, Mathieu Bernard, Juan Benjumea, Xuan-Nga Cao, Lucie Miskic, Charlotte Dugrain, Lucas Ondel, Alan W Black, others
[zero-resource]
Learning weakly supervised multimodal phoneme embeddings
[doi]
[2017] INTERSPEECH 2017: Eighteenth Annual Conference of the International Speech Communication Association.
Rahma Chaabouni, Ewan Dunbar, Neil Zeghidour, Emmanuel Dupoux
The Zero Resource Speech Challenge 2017
[2017] 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU).
Ewan Dunbar, Xuan Nga Cao, Juan Benjumea, Julien Karadayi, Mathieu Bernard, Laurent Besacier, Xavier Anguera, Emmanuel Dupoux
[zero-resource]
Addressing the “two interface” problem: Comparatives and superlatives
[2016] Glossa: a journal of general linguistics.
Ewan Dunbar, Alexis Wellwood
Classification and automatic transcription of primate calls
[doi]
[2016] The Journal of the Acoustical Society of America.
Maarten Versteegh, Jeremy Kuhn, Gabriel Synnaeve, Lucie Ravaux, Emmanuel Chemla, Cristiane Cäsar, James Fuller, Derek Murphy, Anne Schel, Ewan Dunbar
Finite-state phonology predicts a typological gap in cyclic stress assignment
[doi]
[2016] Linguistic Inquiry.
Bronwyn Bjorkman, Ewan Dunbar
Geometric constraints on human speech sound inventories
[doi]
[2016] Frontiers in Psychology: Language Sciences.
Ewan Dunbar, Emmanuel Dupoux
A hybrid dynamic time warping-deep neural network architecture for unsupervised acoustic modeling
[2015] INTERSPEECH 2015: Sixteenth Annual Conference of the International Speech Communication Association.
Roland Thiolliere, Ewan Dunbar, Gabriel Synnaeve, Maarten Versteegh, Emmanuel Dupoux
Quantitative methods for comparing featural representations
[2015] ICPhS 2015: 18th Annual Congress of Phonetic Sciences.
Ewan Dunbar, Gabriel Synnaeve, Emmanuel Dupoux
Self-consistency as an inductive bias in early language acquisition
[2014] Proceedings for the 36th Annual Meeting of the Cognitive Science Society.
Abdellah Fourtassi, Ewan Dunbar, Emmanuel Dupoux
A Bayesian evaluation of the cost of abstractness
[2013] Language Down the Garden Path: The Cognitive and Biological Basis for Linguistic Structures.
Ewan Dunbar, Brian Dillon, William J Idsardi
A single-stage approach to learning phonological categories: Insights from Inuktitut
[doi]
[2013] Cognitive Science.
Brian Dillon, Ewan Dunbar, William Idsardi
[2013] 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.
Aren Jansen, Emmanuel Dupoux, Sharon Goldwater, Mark Johnson, Sanjeev Khudanpur, Kenneth Church, Naomi Feldman, Hynek Hermansky, Florian Metze, Richard Rose, others
The acquisition of phonological inventories
[2012] The Oxford Handbook of Developmental Linguistics.
Ewan Dunbar, William Idsardi
Review of “A critical introduction to phonology: of sound, mind, and body”
[doi]
[2010] Phonology.
Ewan Dunbar, William J Idsardi
Developing intermediate language learning materials: A Labrador Inuttitut story database
[2009] Indigenous Language Revitalization: Encouragement, Guidance & Lessons Learned.
Joan Dicker, Ewan Dunbar, Alana Johns
[2008] Master’s thesis, University of Toronto.
Ewan Dunbar