Daan van Esch
I work on internationalization for language technology at Google, harnessing machine learning and scalable infrastructure to bring support for new languages to products like Gboard and the Assistant. Our world has a wealth of linguistic diversity and it's a fascinating research challenge to build technology across so many different languages.
Authored Publications
Google Publications
Other Publications
Sort By
Now You See Me, Now You Don't: 'Poverty of the Stimulus' Problems and Arbitrary Correspondences in End-to-End Speech Models
Proceedings of the Second Workshop on Computation and Written Language (CAWL) 2024 (to appear)
Preview abstract
End-to-end models for speech recognition and speech synthesis have many benefits, but we argue they also face a unique set of challenges not encountered in conventional multi-stage hybrid systems, which relied on the explicit injection of linguistic knowledge through resources such as phonemic dictionaries and verbalization grammars. These challenges include handling words with unusual grapheme-to-phoneme correspondences, converting between written forms like ‘12’ and spoken forms such as ‘twelve’, and contextual disambiguation of homophones or homographs. We describe the mitigation strategies that have been used for these problems in end-to-end systems, either implicitly or explicitly, and call out that the most commonly used mitigation techniques are likely incompatible with newly emerging approaches that use minimal amounts of supervised audio training data. We review best-of-both-world approaches that allow the use of end-to-end models combined with traditional linguistic resources, which we show are increasingly straightforward to create at scale, and close with an optimistic outlook for bringing speech technologies to many more languages by combining these strands of research.
View details
LinguaMeta: Unified Metadata for Thousands of Languages
Uche Okonkwo
Emily Drummond
Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) (to appear)
Preview abstract
We introduce LinguaMeta, a unified resource for language metadata for thousands of languages, including language codes, names, number of speakers, writing systems, countries, official status, coordinates, and language varieties. The resources are drawn from various existing repositories and supplemented with our own research. Each data point is tagged for its origin, allowing us to easily trace back to and improve existing resources with more up-to-date and complete metadata. The resource is intended for use by researchers and organizations who aim to extend technology to thousands of languages.
View details
Connecting Language Technologies with Rich, Diverse Data Sources Covering Thousands of Languages
Sebastian Ruder
Julia Kreutzer
Clara Rivera
Ishank Saxena
Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) (to appear)
Preview abstract
Contrary to common belief, there are rich and diverse data sources available for many thousands of languages, which can be used to develop technologies for these languages. In this paper, we provide an overview of some of the major online data sources, the types of data that they provide access to, potential applications of this data, and the number of languages that they cover. Even this covers only a small fraction of the data that exists; for example, printed books are published in many languages but few online aggregators exist.
View details
Multimodal Language Identification
Shikhar Bharadwaj
Sriram (Sri) Ganapathy
Sid Dalmia
Wei Han
Yu Zhang
Proceedings of 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024) (2024)
Preview abstract
Spoken language identification refers to the task of automatically predicting the spoken language in a given utterance. Conventionally, it is modeled as a speech-based language identification task. Prior techniques have been constrained to a single modality; however in the case of video data there is a wealth of other metadata that may be beneficial for this task. In this work, we propose MuSeLI, a Multimodal Spoken Language Identification method, which delves into the use of various metadata sources to enhance language identification. Our study reveals that metadata such as video title, description and geographic location provide substantial information to identify the spoken language of the multimedia recording. We conduct experiments using two diverse public datasets of YouTube videos, and obtain state-of-the-art results on the language identification task. We additionally conduct an ablation study that describes the distinct contribution of each modality for language recognition.
View details
Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets
Julia Kreutzer
Lisa Wang
Ahsan Wahab
Nasanbayar Ulzii-Orshikh
Allahsera Auguste Tapo
Nishant Subramani
Artem Sokolov
Claytone Sikasote
Monang Setyawan
Supheakmungkol Sarin
Sokhar Samb
Benoît Sagot
Clara E. Rivera
Annette Rios
Isabel Papadimitriou
Salomey Osei
Pedro Javier Ortiz Suárez
Iroro Fred Ọ̀nọ̀mẹ̀ Orife
Kelechi Ogueji
Rubungo Andre Niyongabo
Toan Nguyen
Mathias Müller
André Müller
Shamsuddeen Hassan Muhammad
Nanda Muhammad
Ayanda Mnyakeni
Jamshidbek Mirzakhalov
Tapiwanashe Matangira
Colin Leong
Nze Lawson
Yacine Jernite
Mathias Jenny
Bonaventure F. P. Dossou
Sakhile Dlamini
Nisansa de Silva
Sakine Çabuk Ballı
Stella Biderman
Alessia Battisti
Ahmed Baruwa
Pallavi Baljekar
Israel Abebe Azime
Ayodele Awokoya
Duygu Ataman
Orevaoghene Ahia
Oghenefego Ahia
Sweta Agrawal
Mofetoluwa Adeyemi
TACL (2022)
Preview abstract
With the success of large-scale pre-training and multilingual modeling in Natural Language Processing (NLP), recent years have seen a proliferation of large, web-mined text datasets covering hundreds of languages. However, to date there has been no systematic analysis of the quality of these publicly available datasets, or whether the datasets actually contain content in the languages they claim to represent. In this work, we manually audit the quality of 205 language-specific corpora released with five major public datasets (CCAligned, ParaCrawl, WikiMatrix, OSCAR, mC4), and audit the correctness of language codes in a sixth (JW300). We find that lower-resource corpora have systematic issues: at least 15 corpora are completely erroneous, and a significant fraction contains less than 50% sentences of acceptable quality. Similarly, we find 82 corpora that are mislabeled or use nonstandard/ambiguous language codes. We demonstrate that these issues are easy to detect even for non-speakers of the languages in question, and supplement the human judgements with automatic analyses. Inspired by our analysis, we recommend techniques to evaluate and improve multilingual corpora and discuss the risks that come with low-quality data releases.
View details
XTREME-S: Evaluating Cross-lingual Speech Representations
Clara E. Rivera
Sebastian Ruder
Simran Khanuja
Ye Jia
Yu Zhang
Proc. Interspeech 2022
Preview abstract
We introduce \xtremes, a new benchmark to evaluate universal cross-lingual speech representations in many languages. XTREME-S covers four task families: speech recognition, classification, retrieval and speech-to-text translation. Covering 102 languages from 10+ language families, 3 different domains and 4 task families, XTREME-S aims to simplify multilingual speech representation evaluation, as well as catalyze research in ``universal'' speech representation learning. This paper describes the new benchmark and establishes the first speech-only and speech-text baselines using XLS-R and mSLAM on all downstream tasks. We motivate the design choices and detail how to use the benchmark. The code and pre-processing scripts will be made publicly available.\footnote{\small\url{https://huggingface.co/datasets/google/xtreme_s}}
View details
Writing System and Speaker Metadata for 2,800+ Language Varieties
Sebastian Ruder
Clara E. Rivera
Proceedings of the Language Resources and Evaluation Conference, European Language Resources Association, Marseille, France (2022), pp. 5035-5046
Preview abstract
We describe an open-source dataset providing metadata for about 2,800 language varieties used in the world today. Specifically, the dataset provides the attested writing system(s) for each of these 2,800+ varieties, as well as an estimated speaker count for each variety. This data set was developed through internal research and has been used for analyses around language technologies. This is the largest publicly-available, machine-readable resource with writing system and speaker information for the world's languages. We hope the availability of this data will catalyze research in under-represented languages.
View details
Handling Compounding in Mobile Keyboard Input
Andreas Christian Kabel
Keith B. Hall
arXiv cs.CL (2022)
Preview abstract
This paper proposes a framework to improve the typing experience of mobile users in morphologically rich languages. Smartphone keyboards typically support features such as input decoding, corrections and predictions that all rely on language models. For latency reasons, these operations happen on device, so the models are of limited size and cannot easily cover all the words needed by users for their daily tasks, especially in morphologically rich languages. In particular, the compounding nature of Germanic languages makes their vocabulary virtually infinite. Similarly, heavily inflecting and agglutinative languages (e.g. Slavic, Turkic or Finno-Ugric languages) tend to have much larger vocabularies than morphologically simpler languages, such as English or Mandarin. We propose to model such languages with automatically selected subword units annotated with what we call binding types, allowing the decoder to know when to bind subword units into words. We show that this method brings around 20% word error rate reduction in a variety of compounding languages. This is more than twice the improvement we previously obtained with a more basic approach, also described in the paper.
View details
Managing Transcription Data for Automatic Speech Recognition with Elpis
Ben Foley
Nay San
The Open Handbook of Linguistic Data Management, The MIT Press (2022)
Preview abstract
This chapter provides a ‘mid-level’ introduction to speech recognition technologies, with particular reference to Elpis (Foley et al., 2018), a tool designed for people with minimal computational experience to take advantage of modern speech recognition technologies in their language documentation transcription workflow. Elpis is intended to be used even in situations where there might not be the large quantities of previously-transcribed recordings typically required for training speech recognition systems. Even in language documentation contexts where people may only have one or two hours of transcribed recordings, using speech recognition can be beneficial to the process of transcription by providing an initial estimate which can be more quickly refined than typed from scratch.
View details
Building Machine Translation Systems for the Next Thousand Languages
Julia Kreutzer
Mengmeng Niu
Pallavi Nikhil Baljekar
Xavier Garcia
Maxim Krikun
Pidong Wang
Apu Shah
Macduff Richard Hughes
Google Research (2022)
Accented Speech Recognition: Benchmarking, Pre-training, and Diverse Data
Zhehuai Chen
Chung-Cheng Chiu
Pavel Golik
Wei Han
Levi King
Suzan Schwartz
(2022)
Preview abstract
Building inclusive speech recognition systems is a crucial step towards developing technologies that speakers of all language varieties can use. Therefore, ASR systems must work for everybody independently of the way they speak. To accomplish this goal, there should be available data sets representing language varieties, and also an understanding of model configuration that is the most helpful in achieving robust understanding of all types of speech. However, there are not enough data sets for accented speech, and for the ones that are already available, more training approaches need to be explored to improve the quality of accented speech recognition. In this paper, we discuss recent progress towards developing more inclusive ASR systems, namely, the importance of building new data sets representing linguistic diversity, and exploring novel training approaches to improve performance for all users. We address recent directions within benchmarking ASR systems for accented speech, measure the effects of wav2vec 2.0 pre-training on accented speech recognition, and highlight corpora relevant for diverse ASR evaluations.
View details
Preview abstract
Almost none of the 2,000+ languages spoken in Africa have widely available automatic speech recognition systems, and the required data is also only available for a few languages. We have experimented with two techniques which may provide pathways to large vocabulary speech recognition for African languages: multilingual modeling and self-supervised learning. We gathered available open source data and collected data for 15 languages, and trained experimental models using these techniques. Our results show that pooling the small amounts of data available in multilingual end-to-end models, and pre-training on unsupervised data can help improve speech recognition quality for many African languages.
View details
How Might We Create Better Benchmarks for Speech Recognition?
James Flynn
Pavel Golik
ACL-IJCNLP 2021 Workshop on Benchmarking: Past, Present and Future (2021)
Preview abstract
The applications of automatic speech recognition (ASR) systems are proliferating, in part due to recent significant quality improvements. However, as recent work indicates, even state-of-the-art speech recognition systems – some which deliver impressive benchmark results, struggle to generalize across use cases. We review relevant work, and, hoping to inform future benchmark development, outline a taxonomy of speech recognition use cases, proposed for the next generation of ASR benchmarks. We also survey work on metrics, in addition to the de facto standard Word Error Rate (WER) metric, and we introduce a versatile framework designed to describe interactions between linguistic variation and ASR performance metrics.
View details
Preview abstract
Pronunciation modeling is a key task for building speech technology in new languages, and while solid grapheme-to-phoneme (G2P) mapping systems exist, language coverage can stand to be improved. The information needed to build G2P models for many more languages can easily be found on Wikipedia, but unfortunately, it is stored in disparate formats. We report on a system we built to mine a pronunciation data set in 819 languages from loosely structured tables within Wikipedia. The data includes phoneme inventories, and for 63 low-resource languages, also includes the grapheme-to-phoneme (G2P) mapping. 54 of these languages do not have easily findable G2P mappings online otherwise. We turned the information from Wikipedia into a structured, machine-readable TSV format, and make the resulting data set publicly available so it can be improved further and used in a variety of applications involving low-resource languages.
View details
Preview abstract
Large text corpora are increasingly important for a wide variety of Natural Language Processing (NLP) tasks, and automatic language identification (LangID) is a core technology needed to collect such datasets in a multilingual context. LangID is largely treated as solved in the literature, with models reported that achieve over 90% average F1 on as many as 1,366 languages. We train LangID models on up to 1,629 languages with comparable quality on held-out test sets, but find that human-judged LangID accuracy for web-crawl text corpora created using these models is only around 5% for many lower-resource languages, suggesting a need for more robust evaluation. Further analysis revealed a variety of error modes, arising from domain mismatch, class imbalance, language similarity, and insufficiently expressive models. We propose two classes of techniques to mitigate these errors: wordlist-based tunable-precision filters (for which we release curated lists in about 500 languages) and transformer-based semi-supervised LangID models, which increase median dataset precision from 5.5% to 71.2%. These techniques enable us to create an initial data set covering 100K or more relatively clean sentences in each of 500+ languages, paving the way towards a 1,000-language web text corpus.
View details
Data-Driven Parametric Text Normalization: Rapidly Scaling Finite-State Transduction Verbalizers to New Languages
Kim Anne Heiligenstein
Nikos Bampounis
Christian Schallhart
Jonas Fromseier Mortensen
Proceedings of the 1st Joint SLTU and CCURL Workshop (SLTU-CCURL 2020), Language Resources and Evaluation Conference (LREC 2020), Marseille, 218–225
Preview abstract
This paper presents a methodology for rapidly generating FST-based verbalizers for ASR and TTS systems by efficiently sourcing language-specific data. We describe a questionnaire which collects the necessary data to bootstrap the number grammar induction system and parameterize the verbalizer templates described in Ritchie et al. (2019), and a machine-readable data store which allows the data collected through the questionnaire to be supplemented by additional data from other sources. We also discuss the benefits of this system for low-resource languages.
View details
Future Directions in Technological Support for Language Documentation
Ben Foley
Nay San
Proceedings of the 3rd Workshop on the Use of Computational Methods in the Study of Endangered Languages (ComputEL-3) (2019)
Preview abstract
To reduce the annotation burden placed on linguistic fieldworkers, freeing up time for deeper linguistic analysis and descriptive work, the language documentation community has been working with machine learning researchers to investigate what role technology can play, with promising early results. This paper describes a number of potential follow-up technical projects that we believe would be worthwhile and straightforward to do. We provide examples of the annotation tasks for computer scientists; descriptions of the technological challenges involved and the estimated level of complexity; and pointers to relevant literature. We hope providing a clear overview of what the needs are and what annotation challenges exist will help facilitate the dialogue and collaboration between computer scientists and fieldwork linguists.
View details
Unified Verbalization for Speech Recognition & Synthesis Across Languages
Christian Schallhart
Nikos Bampounis
Jonas Fromseier Mortensen
Millie Holt
Proceedings of Interspeech 2019
Preview abstract
We describe a new approach to converting written tokens to their spoken form, which can be used across automatic speech recognition (ASR) and text-to-speech synthesis (TTS) systems. Both ASR and TTS systems need to map from the written to the spoken domain, and we present an approach that enables us to share verbalization grammars between the two systems. We also describe improvements to an induction system for number name grammars. Between these shared ASR/TTS verbalization systems and the improved induction system for number name grammars, we see significant gains in development time and scalability across languages
View details
Preview abstract
We present our approach to automatically generating keyboard layouts on mobile devices for typing low-resource languages written in the Latin script. For many speakers, one of the barriers in accessing and creating text content on the web is the absence of input tools for their language. Ease in typing in these languages would lower technological barriers to online communication and collaboration, likely leading to the creation of more web content. Unfortunately, it can be time-consuming to develop layouts manually even for language communities that use a keyboard layout very similar to English: starting from scratch requires many configuration files to describe multiple possible behaviors for each key. With our approach, we only need a small amount of data in each language to generate keyboard layouts with very little human effort. This process can help serve speakers of low-resource languages in a scalable way, allowing us to develop input tools for more languages. Having input tools that reflect the linguistic diversity of the world will let as many people as possible use technology to learn, communicate, and share their thoughts in their own native languages.
View details
Writing Across the World's Languages: Deep Internationalization for Gboard, the Google Keyboard
Elnaz Sarbar
Jeremy O'Brien
Evan Elizabeth Crew
Chieu Nguyen
arXiv cs.HC (2019)
Preview abstract
This technical report describes our deep internationalization program for Gboard, the Google Keyboard. Today, Gboard supports 900+ language varieties across 70+ writing systems, and this report describes how and why we added support for these language varieties from around the globe. Many languages of the world are increasingly used in writing on an everyday basis, and we describe the trends we see. We cover technological and logistical challenges in scaling up a language technology product like Gboard to hundreds of language varieties, and describe how we built systems and processes to operate at scale. Finally, we summarize the key take-aways from user studies we ran with speakers of hundreds of languages from around the world.
View details
Preview abstract
We discuss two methods that let us easily create grapheme-to-phoneme (G2P) conversion systems for languages without any human-curated pronunciation lexicons, as long as we know the phoneme inventory of the target language and as long as we have some pronunciation lexicons for other languages written in the same script. We use these resources to infer what grapheme-to-phoneme correspondences we would expect, and predict pronunciations for words in the target language with minimal or no language-specific human work. Our first approach uses finite-state transducers, while our second approach uses a sequence-to-sequence neural network. Our G2P models reach high degrees of accuracy, and can be used for various applications, e.g. in developing an Automatic Speech Recognition system. Our methods greatly simplify a task that has historically required extensive manual labor.
View details
Preview abstract
When building automatic speech recognition (ASR) systems, typically some amount of audio and text data in the target language is needed. While text data can be obtained relatively easily across many languages, transcribed audio data is challenging to obtain. This presents a barrier to making voice technologies available in more languages of the world. In this paper, we present a way to build an ASR system for a language even in the absence of any audio training data in that language at all. We do this by simply re-using an existing acoustic model from a phonologically similar language, without any kind of modification or adaptation towards the target language. The basic insight is that, if two languages are sufficiently similar in terms of their phonological system, an acoustic model should hold up relatively well when used for another language. We describe how we tailor our pronunciation models to enable such re-use, and show experimental results across a number of languages from various language families. We also provide a theoretical analysis of situations in which this approach is likely to work. Our results show that is possible to achieve less than 20% word error rate (WER) using this method.
View details
Text Normalization Infrastructure that Scales to Hundreds of Language Varieties
Mason Chua
Noah Coccaro
Eunjoon Cho
Sujeet Bhandari
Libin Jia
Proceedings of the 11th edition of the Language Resources and Evaluation Conference (2018)
Preview abstract
We describe the automated multi-language text normalization infrastructure that prepares textual data to train language models used in Google's keyboards and speech recognition systems, across hundreds of language varieties. Training corpora are sourced from various types of data sets, and the text is then normalized using a sequence of hand-written grammars and learned models. These systems need to scale to hundreds or thousands of language varieties in order to meet product needs. Frequent data refreshes, privacy considerations and simultaneous updates across such a high number of languages make manual inspection of the normalized training data infeasible, while there is ample opportunity for data normalization issues. By tracking metrics about the data and how it was processed, we are able to catch internal data processing issues and external data corruption issues that can be hard to notice using standard extrinsic evaluation methods. Showing the importance of paying attention to data normalization behavior in large-scale pipelines, these metrics have highlighted issues in Google's real-world speech recognition system that have caused significant, but latent, quality degradation.
View details
Building Speech Recognition Systems for Language Documentation: The CoEDL Endangered Language Pipeline and Inference System
Ben Foley
Josh Arnold
Rolando Coto-Solano
Gautier Durantin
T. Mark Ellison
Scott Heath
František Kratochvíl
Zara Maxwell-Smith
David Nash
Ola Olsson
Mark Richards
Nay San
Hywel Stoakes
Nick Thieberger
Janet Wiles
Proceedings of the 6th International Workshop on Spoken Language Technologies for Under-resourced Languages (SLTU 2018)
Preview abstract
Machine learning has revolutionised speech technologies for major world languages, but these technologies have generally not been available for the roughly 4,000 languages with populations of fewer than 10,000 speakers. This paper describes the development of Elpis, a pipeline which language documentation workers with minimal computational experience can use to build their own speech recognition models, resulting in models being built for 16 languages from the Asia-Pacific region. Elpis puts machine learning speech technologies within reach of people working with languages with scarce data, in a scalable way. This is impactful since it enables language communities to cross the digital divide, and speeds up language documentation. Complete automation of the process is not feasible for languages with small quantities of data and potentially large vocabularies. Hence our goal is not full automation, but rather to make a practical and effective workflow that integrates machine learning technologies.
View details
Mining Training Data for Language Modeling across the World’s Languages
Proceedings of the 6th International Workshop on Spoken Language Technologies for Under-resourced Languages (SLTU 2018)
Preview abstract
Building smart keyboards and speech recognition systems for new languages requires a large, clean text corpus to train n-gram language models on. We report our findings on how much text data can realistically be found on the web across thousands of languages. In addition, we describe an innovative, scalable approach to normalizing this data: all data sources are noisy to some extent, but this situation is even more severe for low-resource languages. To help clean the data we find across all languages in a scalable way, we built a pipeline to automatically derive the configuration for language-specific text normalization systems, which we describe here as well.
View details
Preview abstract
We describe an expanded taxonomy of semiotic classes for text normalization, building upon the work in Sproat (2001). We add a large number of categories of non-standard words (NSWs) that we believe a robust real-world text normalization system will have to be able to process. Our new categories are based upon empirical findings encountered while building text normalization systems across many languages, for both Speech Recognition and Speech Synthesis purposes. We believe our new taxonomy is useful both for ensuring high coverage when writing manual grammars, as well as for eliciting training data to build machine learning-based text normalization systems.
View details
Preview abstract
Word pronunciations, consisting of phoneme sequences and the associated syllabification and stress patterns, are vital for both speech recognition and text-to-speech (TTS) systems. For speech recognition phoneme sequences for words may be learned from audio data. We train recurrent neural network (RNN) based models to predict the syllabification and stress pattern for such pronunciations making them usable for TTS. We find these RNN models significantly outperform naive rulebased models for almost all languages we tested. Further, we find additional improvements to the stress prediction model by using the spelling as features in addition to the phoneme sequence. Finally, we train a single RNN model to predict the phoneme sequence, syllabification and stress for a given word. For several languages, this single RNN outperforms similar models trained specifically for either phoneme sequence or stress prediction. We report an exhaustive comparison of these approaches for twenty languages.
View details
No Results Found