Alert button
Picture for Min Ma

Min Ma

Alert button

Finite-Time Adaptive Fuzzy Tracking Control for Nonlinear State Constrained Pure-Feedback Systems

Oct 23, 2023
Ju Wu, Tong Wang, Min Ma

This paper investigates the finite-time adaptive fuzzy tracking control problem for a class of pure-feedback system with full-state constraints. With the help of Mean-Value Theorem, the pure-feedback nonlinear system is transformed into strict-feedback case. By employing finite-time-stable like function and state transformation for output tracking error, the output tracking error converges to a predefined set in a fixed finite interval. To tackle the problem of state constraints, integral Barrier Lyapunov functions are utilized to guarantee that the state variables remain within the prescribed constraints with feasibility check. Fuzzy logic systems are utilized to approximate the unknown nonlinear functions. In addition, all the signals in the closed-loop system are guaranteed to be semi-global ultimately uniformly bounded. Finally, two simulation examples are given to show the effectiveness of the proposed control strategy.

Viaarxiv icon

Multimodal Modeling For Spoken Language Identification

Sep 19, 2023
Shikhar Bharadwaj, Min Ma, Shikhar Vashishth, Ankur Bapna, Sriram Ganapathy, Vera Axelrod, Siddharth Dalmia, Wei Han, Yu Zhang, Daan van Esch, Sandy Ritchie, Partha Talukdar, Jason Riesa

Figure 1 for Multimodal Modeling For Spoken Language Identification
Figure 2 for Multimodal Modeling For Spoken Language Identification
Figure 3 for Multimodal Modeling For Spoken Language Identification
Figure 4 for Multimodal Modeling For Spoken Language Identification

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.

Viaarxiv icon

MASR: Metadata Aware Speech Representation

Jul 20, 2023
Anjali Raj, Shikhar Bharadwaj, Sriram Ganapathy, Min Ma, Shikhar Vashishth

Figure 1 for MASR: Metadata Aware Speech Representation
Figure 2 for MASR: Metadata Aware Speech Representation
Figure 3 for MASR: Metadata Aware Speech Representation
Figure 4 for MASR: Metadata Aware Speech Representation

In the recent years, speech representation learning is constructed primarily as a self-supervised learning (SSL) task, using the raw audio signal alone, while ignoring the side-information that is often available for a given speech recording. In this paper, we propose MASR, a Metadata Aware Speech Representation learning framework, which addresses the aforementioned limitations. MASR enables the inclusion of multiple external knowledge sources to enhance the utilization of meta-data information. The external knowledge sources are incorporated in the form of sample-level pair-wise similarity matrices that are useful in a hard-mining loss. A key advantage of the MASR framework is that it can be combined with any choice of SSL method. Using MASR representations, we perform evaluations on several downstream tasks such as language identification, speech recognition and other non-semantic tasks such as speaker and emotion recognition. In these experiments, we illustrate significant performance improvements for the MASR over other established benchmarks. We perform a detailed analysis on the language identification task to provide insights on how the proposed loss function enables the representations to separate closely related languages.

Viaarxiv icon

Label Aware Speech Representation Learning For Language Identification

Jun 07, 2023
Shikhar Vashishth, Shikhar Bharadwaj, Sriram Ganapathy, Ankur Bapna, Min Ma, Wei Han, Vera Axelrod, Partha Talukdar

Figure 1 for Label Aware Speech Representation Learning For Language Identification
Figure 2 for Label Aware Speech Representation Learning For Language Identification
Figure 3 for Label Aware Speech Representation Learning For Language Identification
Figure 4 for Label Aware Speech Representation Learning For Language Identification

Speech representation learning approaches for non-semantic tasks such as language recognition have either explored supervised embedding extraction methods using a classifier model or self-supervised representation learning approaches using raw data. In this paper, we propose a novel framework of combining self-supervised representation learning with the language label information for the pre-training task. This framework, termed as Label Aware Speech Representation (LASR) learning, uses a triplet based objective function to incorporate language labels along with the self-supervised loss function. The speech representations are further fine-tuned for the downstream task. The language recognition experiments are performed on two public datasets - FLEURS and Dhwani. In these experiments, we illustrate that the proposed LASR framework improves over the state-of-the-art systems on language identification. We also report an analysis of the robustness of LASR approach to noisy/missing labels as well as its application to multi-lingual speech recognition tasks.

* Accepted at Interspeech 2023 
Viaarxiv icon

XTREME-UP: A User-Centric Scarce-Data Benchmark for Under-Represented Languages

May 24, 2023
Sebastian Ruder, Jonathan H. Clark, Alexander Gutkin, Mihir Kale, Min Ma, Massimo Nicosia, Shruti Rijhwani, Parker Riley, Jean-Michel A. Sarr, Xinyi Wang, John Wieting, Nitish Gupta, Anna Katanova, Christo Kirov, Dana L. Dickinson, Brian Roark, Bidisha Samanta, Connie Tao, David I. Adelani, Vera Axelrod, Isaac Caswell, Colin Cherry, Dan Garrette, Reeve Ingle, Melvin Johnson, Dmitry Panteleev, Partha Talukdar

Figure 1 for XTREME-UP: A User-Centric Scarce-Data Benchmark for Under-Represented Languages
Figure 2 for XTREME-UP: A User-Centric Scarce-Data Benchmark for Under-Represented Languages
Figure 3 for XTREME-UP: A User-Centric Scarce-Data Benchmark for Under-Represented Languages
Figure 4 for XTREME-UP: A User-Centric Scarce-Data Benchmark for Under-Represented Languages

Data scarcity is a crucial issue for the development of highly multilingual NLP systems. Yet for many under-represented languages (ULs) -- languages for which NLP re-search is particularly far behind in meeting user needs -- it is feasible to annotate small amounts of data. Motivated by this, we propose XTREME-UP, a benchmark defined by: its focus on the scarce-data scenario rather than zero-shot; its focus on user-centric tasks -- tasks with broad adoption by speakers of high-resource languages; and its focus on under-represented languages where this scarce-data scenario tends to be most realistic. XTREME-UP evaluates the capabilities of language models across 88 under-represented languages over 9 key user-centric technologies including ASR, OCR, MT, and information access tasks that are of general utility. We create new datasets for OCR, autocomplete, semantic parsing, and transliteration, and build on and refine existing datasets for other tasks. XTREME-UP provides methodology for evaluating many modeling scenarios including text-only, multi-modal (vision, audio, and text),supervised parameter tuning, and in-context learning. We evaluate commonly used models on the benchmark. We release all code and scripts to train and evaluate models

Viaarxiv icon

FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech

May 25, 2022
Alexis Conneau, Min Ma, Simran Khanuja, Yu Zhang, Vera Axelrod, Siddharth Dalmia, Jason Riesa, Clara Rivera, Ankur Bapna

Figure 1 for FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech
Figure 2 for FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech
Figure 3 for FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech
Figure 4 for FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech

We introduce FLEURS, the Few-shot Learning Evaluation of Universal Representations of Speech benchmark. FLEURS is an n-way parallel speech dataset in 102 languages built on top of the machine translation FLoRes-101 benchmark, with approximately 12 hours of speech supervision per language. FLEURS can be used for a variety of speech tasks, including Automatic Speech Recognition (ASR), Speech Language Identification (Speech LangID), Translation and Retrieval. In this paper, we provide baselines for the tasks based on multilingual pre-trained models like mSLAM. The goal of FLEURS is to enable speech technology in more languages and catalyze research in low-resource speech understanding.

Viaarxiv icon

XTREME-S: Evaluating Cross-lingual Speech Representations

Apr 13, 2022
Alexis Conneau, Ankur Bapna, Yu Zhang, Min Ma, Patrick von Platen, Anton Lozhkov, Colin Cherry, Ye Jia, Clara Rivera, Mihir Kale, Daan Van Esch, Vera Axelrod, Simran Khanuja, Jonathan H. Clark, Orhan Firat, Michael Auli, Sebastian Ruder, Jason Riesa, Melvin Johnson

Figure 1 for XTREME-S: Evaluating Cross-lingual Speech Representations
Figure 2 for XTREME-S: Evaluating Cross-lingual Speech Representations
Figure 3 for XTREME-S: Evaluating Cross-lingual Speech Representations
Figure 4 for XTREME-S: Evaluating Cross-lingual Speech Representations

We introduce XTREME-S, a new benchmark to evaluate universal cross-lingual speech representations in many languages. XTREME-S covers four task families: speech recognition, classification, speech-to-text translation and retrieval. 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. Datasets and fine-tuning scripts are made easily accessible at https://hf.co/datasets/google/xtreme_s.

* Minor fix: language code for Filipino (Tagalog), "tg" -> "tl" 
Viaarxiv icon

BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition

Oct 01, 2021
Yu Zhang, Daniel S. Park, Wei Han, James Qin, Anmol Gulati, Joel Shor, Aren Jansen, Yuanzhong Xu, Yanping Huang, Shibo Wang, Zongwei Zhou, Bo Li, Min Ma, William Chan, Jiahui Yu, Yongqiang Wang, Liangliang Cao, Khe Chai Sim, Bhuvana Ramabhadran, Tara N. Sainath, Françoise Beaufays, Zhifeng Chen, Quoc V. Le, Chung-Cheng Chiu, Ruoming Pang, Yonghui Wu

Figure 1 for BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition
Figure 2 for BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition
Figure 3 for BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition
Figure 4 for BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition

We summarize the results of a host of efforts using giant automatic speech recognition (ASR) models pre-trained using large, diverse unlabeled datasets containing approximately a million hours of audio. We find that the combination of pre-training, self-training and scaling up model size greatly increases data efficiency, even for extremely large tasks with tens of thousands of hours of labeled data. In particular, on an ASR task with 34k hours of labeled data, by fine-tuning an 8 billion parameter pre-trained Conformer model we can match state-of-the-art (SoTA) performance with only 3% of the training data and significantly improve SoTA with the full training set. We also report on the universal benefits gained from using big pre-trained and self-trained models for a large set of downstream tasks that cover a wide range of speech domains and span multiple orders of magnitudes of dataset sizes, including obtaining SoTA performance on many public benchmarks. In addition, we utilize the learned representation of pre-trained networks to achieve SoTA results on non-ASR tasks.

* 14 pages, 7 figures, 13 tables; v2: minor corrections, reference baselines and bibliography updated 
Viaarxiv icon

Scaling End-to-End Models for Large-Scale Multilingual ASR

Apr 30, 2021
Bo Li, Ruoming Pang, Tara N. Sainath, Anmol Gulati, Yu Zhang, James Qin, Parisa Haghani, W. Ronny Huang, Min Ma

Figure 1 for Scaling End-to-End Models for Large-Scale Multilingual ASR
Figure 2 for Scaling End-to-End Models for Large-Scale Multilingual ASR
Figure 3 for Scaling End-to-End Models for Large-Scale Multilingual ASR

Building ASR models across many language families is a challenging multi-task learning problem due to large language variations and heavily unbalanced data. Existing work has shown positive transfer from high resource to low resource languages. However, degradations on high resource languages are commonly observed due to interference from the heterogeneous multilingual data and reduction in per-language capacity. We conduct a capacity study on a 15-language task, with the amount of data per language varying from 7.7K to 54.7K hours. We adopt GShard [1] to efficiently scale up to 10B parameters. Empirically, we find that (1) scaling the number of model parameters is an effective way to solve the capacity bottleneck - our 500M-param model is already better than monolingual baselines and scaling it to 1B and 10B brought further quality gains; (2) larger models are not only more data efficient, but also more efficient in terms of training cost as measured in TPU days - the 1B-param model reaches the same accuracy at 34% of training time as the 500M-param model; (3) given a fixed capacity budget, adding depth usually works better than width and large encoders tend to do better than large decoders.

* Submitted to INTERSPEECH 2021 
Viaarxiv icon

Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data

Oct 22, 2020
Thibault Doutre, Wei Han, Min Ma, Zhiyun Lu, Chung-Cheng Chiu, Ruoming Pang, Arun Narayanan, Ananya Misra, Yu Zhang, Liangliang Cao

Figure 1 for Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data
Figure 2 for Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data
Figure 3 for Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data
Figure 4 for Improving Streaming Automatic Speech Recognition With Non-Streaming Model Distillation On Unsupervised Data

Streaming end-to-end automatic speech recognition (ASR) models are widely used on smart speakers and on-device applications. Since these models are expected to transcribe speech with minimal latency, they are constrained to be causal with no future context, compared to their non-streaming counterparts. Consequently, streaming models usually perform worse than non-streaming models. We propose a novel and effective learning method by leveraging a non-streaming ASR model as a teacher to generate transcripts on an arbitrarily large data set, which is then used to distill knowledge into streaming ASR models. This way, we scale the training of streaming models to up to 3 million hours of YouTube audio. Experiments show that our approach can significantly reduce the word error rate (WER) of RNNT models not only on LibriSpeech but also on YouTube data in four languages. For example, in French, we are able to reduce the WER by 16.4% relatively to a baseline streaming model by leveraging a non-streaming teacher model trained on the same amount of labeled data as the baseline.

Viaarxiv icon