Titouan Parcollet

University of Avignon | UAPV · Laboratoire Informatique d'Avignon (EA 4128)

Automatic speech recognition (ASR) with an encoder equipped with self-attention, whether streaming or non-streaming, takes quadratic time in the length of the speech utterance. This slows down training and decoding, increase their cost, and limit the deployment of the ASR in constrained devices. SummaryMixing is a promising linear-time complexity a...

Self-Supervised Learning (SSL) has proven to be effective in various domains, including speech processing. However, SSL is computationally and memory expensive. This is in part due the quadratic complexity of multi-head self-attention (MHSA). Alternatives for MHSA have been proposed and used in the speech domain, but have yet to be investigated pro...

Self-supervised learning (SSL) models usually require weeks of pre-training with dozens of high-end GPUs. These models typically have a multi-headed self-attention (MHSA) context encoder. However, MHSA takes quadratic time and space in the input length, contributing to the high pre-training cost. Linear-complexity alternatives to MHSA have been pro...

Recent advancements in textless speech-to-speech translation systems have been driven by the adoption of self-supervised learning techniques. Although most state-of-the-art systems adopt a similar architecture to transform source language speech into sequences of discrete representations in the target language, the criteria for selecting these targ...

Despite being trained on massive and diverse datasets, speech self-supervised encoders are generally used for downstream purposes as mere frozen feature extractors or model initializers before fine-tuning. The former severely limits the exploitation of large encoders, while the latter hurts the robustness acquired during pretraining, especially in...

SpeechBrain is an open-source Conversational AI toolkit based on PyTorch, focused particularly on speech processing tasks such as speech recognition, speech enhancement, speaker recognition, text-to-speech, and much more.It promotes transparency and replicability by releasing both the pre-trained models and the complete "recipes" of code and algori...

Textless speech-to-speech translation systems are rapidly advancing, thanks to the integration of self-supervised learning techniques. However, existing state-of-the-art systems fall short when it comes to capturing and transferring expressivity accurately across different languages. Expressivity plays a vital role in conveying emotions, nuances, a...

Self-supervised learning (SSL) leverages large datasets of unlabeled speech to reach impressive performance with reduced amounts of annotated data. The high number of proposed approaches fostered the emergence of comprehensive benchmarks that evaluate their performance on a set of downstream tasks exploring various aspects of the speech signal. How...

State-of-the-art ASR systems have achieved promising results by modeling local and global interactions separately. While the former can be computed efficiently, global interactions are usually modeled via attention mechanisms, which are expensive for long input sequences. Here, we address this by extending HyperMixer, an efficient alternative to at...

Modern speech recognition systems rely on self-attention. Unfortunately, token mixing with self-attention takes quadratic time in the length of the speech utterance, slowing down inference as well as training and increasing memory consumption. Cheaper alternatives to self-attention for ASR have been developed, but fail to consistently reach the sam...

We propose a method for speech-to-speech emotionpreserving translation that operates at the level of discrete speech units. Our approach relies on the use of multilingual emotion embedding that can capture affective information in a language-independent manner. We show that this embedding can be used to predict the pitch and duration of speech unit...

State-of-the-art ASR systems have achieved promising results by modeling local and global interactions separately. While the former can be computed efficiently, global interactions are usually modeled via attention mechanisms, which are expensive for long input sequences. Here, we address this by extending HyperMixer, an efficient alternative to at...

Self-supervised learning (SSL) has allowed substantial progress in Automatic Speech Recognition (ASR) performance in low-resource settings. In this context, it has been demonstrated that larger self-supervised feature extractors are crucial for achieving lower downstream ASR error rates. Thus, better performance might be sanctioned with longer infe...

The light gated recurrent units (Li-GRU) is well-known for achieving impressive results in automatic speech recognition (ASR) tasks while being lighter and faster to train than a standard gated recurrent units (GRU). However, the unbounded nature of its rectified linear unit on the candidate recurrent gate induces an important gradient exploding ph...

Through solving pretext tasks, self-supervised learning leverages unlabeled data to extract useful latent representations replacing traditional input features in the downstream task. In audio and speech signal processing, a wide range of features were engineered through decades of research efforts. As it turns out, learning to predict such features...

Self-supervised learning (SSL) has proven vital in speech and audio-related applications. The paradigm trains a general model on unlabeled data that can later be used to solve specific downstream tasks. This type of model is costly to train as it requires manipulating long input sequences that can only be handled by powerful centralised servers. Su...

When the available hardware cannot meet the memory and compute requirements to efficiently train high performing machine learning models, a compromise in either the training quality or the model complexity is needed. In Federated Learning (FL), nodes are orders of magnitude more constrained than traditional server-grade hardware and are often batte...

L'apprentissage auto-supervisé a ouvert des perspectives prometteuses dans de nombreux domaines comme la vision par ordinateur, le traitement automatique de la langue ou celui de la parole. Les modèles pré-appris sur de grandes quantités de données non étiquetées peuvent être ajustés sur de petits ensembles de données transcrites manuellement. Ceux...

L'apprentissage autosupervisé a apporté des améliorations remarquables dans de nombreux domaines tels que la vision par ordinateur ou le traitement de la langue et de la parole, en exploitant de grandes quantités de données non étiquetées. Dans le contexte spécifique de la parole, cependant, et malgré des résultats prometteurs, il existe un manque...

Training Automatic Speech Recognition (ASR) models under federated learning (FL) settings has attracted a lot of attention recently. However, the FL scenarios often presented in the literature are artificial and fail to capture the complexity of real FL systems. In this paper, we construct a challenging and realistic ASR federated experimental setu...

Contrastive learning enables learning useful audio and speech representations without ground-truth labels by maximizing the similarity between latent representations of similar signal segments. In this framework various data augmentation techniques are usually exploited to help enforce desired invariances within the learned representations, improvi...

The ubiquity of microphone-enabled devices has lead to large amounts of unlabelled audio data being produced at the edge. The integration of self-supervised learning (SSL) and federated learning (FL) into one coherent system can potentially offer data privacy guarantees while also advancing the quality and robustness of speech representations. In t...

Recent works showed that end-to-end neural approaches tend to become very popular for spoken language understanding (SLU). Through the term end-to-end, one considers the use of a single model optimized to extract semantic information directly from the speech signal. A major issue for such models is the lack of paired audio and textual data with sem...

In speech technologies, speaker's voice representation is used in many applications such as speech recognition, voice conversion , speech synthesis and, obviously, user authentication. Modern vocal representations of the speaker are based on neural embeddings. In addition to the targeted information, these representations usually contain sensitive...

Through solving pretext tasks, self-supervised learning leverages unlabeled data to extract useful latent representations replacing traditional input features in the downstream task. In various application domains, including computer vision, natural language processing and audio/speech signal processing, a wide range of features where engineered th...

SpeechBrain is an open-source and all-in-one speech toolkit. It is designed to facilitate the research and development of neural speech processing technologies by being simple, flexible, user-friendly, and well-documented. This paper describes the core architecture designed to support several tasks of common interest, allowing users to naturally co...

Training Automatic Speech Recognition (ASR) models under federated learning (FL) settings has recently attracted considerable attention. However, the FL scenarios often presented in the literature are artificial and fail to capture the complexity of real FL systems. In this paper, we construct a challenging and realistic ASR federated experimental...

Self-Supervised Learning (SSL) using huge unlabeled data has been successfully explored for image and natural language processing. Recent works also investigated SSL from speech. They were notably successful to improve performance on downstream tasks such as automatic speech recognition (ASR). While these works suggest it is possible to reduce depe...

Through solving pretext tasks, self-supervised learning (SSL) leverages unlabeled data to extract useful latent representations replacing traditional input features in the downstream task. A common pretext task consists in pretraining a SSL model on pseudo-labels derived from the original signal. This technique is particularly relevant for speech d...

Federated Learning (FL) allows edge devices to collaboratively learn a shared prediction model while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store data in the cloud. Despite the algorithmic advancements in FL, the support for on-device training of FL algorithms on edge device...

This paper introduces Timers and Such, a new open source dataset of spoken English commands for common voice control use cases involving numbers. We describe the gap in existing spoken language understanding datasets that Timers and Such fills, the design and creation of the dataset, and experiments with a number of ASR-based and end-to-end baselin...

Despite impressive results, deep learning-based technologies also raise severe privacy and environmental concerns induced by the training procedure often conducted in datacenters. In response, alternatives to centralized training such as Federated Learning (FL) have emerged. Perhaps unexpectedly, FL, in particular, is starting to be deployed at a g...

Despite impressive results, deep learning-based technologies also raise severe privacy and environmental concerns induced by the training procedure often conducted in data centers. In response, alternatives to centralized training such as Federated Learning (FL) have emerged. Perhaps unexpectedly, FL, in particular, is starting to be deployed at a...

Despite the significant progress in automatic speech recognition (ASR), distant ASR remains challenging due to noise and reverberation. A common approach to mitigate this issue consists of equipping the recording devices with multiple microphones that capture the acoustic scene from different perspectives. These multi-channel audio recordings conta...

Despite impressive results, deep learning-based technologies also raise severe privacy and environmental concerns induced by the training procedure often conducted in data centers. In response, alternatives to centralized training such as Federated Learning (FL) have emerged. Perhaps unexpectedly, FL in particular is starting to be deployed at a gl...

With the increasing interest over speech technologies, numerous Automatic Speaker Verification (ASV) systems are employed to perform person identification. In the latter context, the systems rely on neural embeddings as a speaker representation. Nonetheless, such representations may contain privacy sensitive information about the speakers (e.g. age...

Federated Learning (FL) has emerged as a promising technique for edge devices to collaboratively learn a shared prediction model, while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store the data in the cloud. However, FL is difficult to implement and deploy in practice, consideri...

Knowledge distillation has been widely used to compress existing deep learning models while preserving the performance on a wide range of applications. In the specific context of Automatic Speech Recognition (ASR), distillation from ensembles of acoustic models has recently shown promising results in increasing recognition performance. In this pape...

Distant speech recognition remains a challenging application for modern deep learning based Automatic Speech Recognition (ASR) systems, due to complex recording conditions involving noise and reverberation. Multiple microphones are commonly combined with well-known speech processing techniques to enhance the original signals and thus enhance the sp...

Despite the significant progress in automatic speech recognition (ASR), distant ASR remains challenging due to noise and reverberation. A common approach to mitigate this issue consists of equipping the recording devices with multiple microphones that capture the acoustic scene from different perspectives. These multi-channel audio recordings conta...

Quaternion neural networks have recently received an increasing interest due to noticeable improvements over real-valued neural networks on real world tasks such as image, speech and signal processing. The extension of quaternion numbers to neural architectures reached state-of-the-art performances with a reduction of the number of neural parameter...

Convolutional Neural Networks (CNN) have been used in Automatic Speech Recognition (ASR) to learn representations directly from the raw signal instead of hand-crafted acoustic features, providing a richer and lossless input signal. Recent researches propose to inject prior acoustic knowledge to the first convolutional layer by integrating the shape...

In the recent years, deep learning has become the leading approach to modern artificial intelligence (AI). The important improvement in terms of processing time required for learning AI based models alongside with the growing amount of available data made of deep neural networks (DNN) the strongest solution to solve complex real-world problems. How...

Modern end-to-end (E2E) Automatic Speech Recognition (ASR) systems rely on Deep Neural Networks (DNN) that are mostly trained on handcrafted and pre-computed acoustic features such as Mel-filter-banks or Mel-frequency cepstral coefficients. Nonetheless , and despite worse performances, E2E ASR models processing raw waveforms are an active research...

Machine learning (ML) and deep learning with deep neural networks (DNN), have drastically improved the performances of modern systems on numerous spoken language understanding (SLU) related tasks. Since most of current researches focus on new neural architectures to enhance the performances in realistic conditions, few recent works investigated the...

Deep learning is at the core of recent spoken language understanding (SLU) related tasks. More precisely, deep neu-ral networks (DNNs) drastically increased the performances of SLU systems, and numerous architectures have been proposed. In the real-life context of theme identification of telephone conversations , it is common to hold both a human,...

Deep neural networks (DNNs) and more precisely recurrent neural networks (RNNs) are at the core of modern automatic speech recognition systems, due to their efficiency to process input sequences. Recently, it has been shown that different input representations, based on multidimensional algebras, such as complex and quaternion numbers, are able to...

Conversely to the legislation that struggles to develop, regulate and supervise the use of artificial intelligence (AI), the civil society, that gradually realizes the fundamental issues and perspectives induced by this new technology, slowly starts to take responsibility and to mobilize. Social responsibility expresses itself through the emergence...

Recurrent neural networks (RNNs) are powerful architectures to model sequential data, due to their capability to learn short and long-term dependencies between the basic elements of a sequence. Nonetheless, popular tasks such as speech or images recognition, involve multi-dimensional input features that are characterized by strong internal dependen...

Deep learning is at the core of recent spoken language understanding (SLU) related tasks. More precisely, deep neu-ral networks (DNNs) drastically increased the performances of SLU systems, and numerous architectures have been proposed. In the real-life context of theme identification of telephone conversations , it is common to hold both a human,...

Neural network architectures are at the core of powerful automatic speech recognition systems (ASR). However, while recent researches focus on novel model architectures, the acoustic input features remain almost unchanged. Traditional ASR systems rely on multidimensional acoustic features such as the Mel filter bank energies alongside with the firs...

Neural network architectures are at the core of powerful automatic speech recognition systems (ASR). However, while recent researches focus on novel model architectures, the acoustic input features remain almost unchanged. Traditional ASR systems rely on multidimensional acoustic features such as the Mel filter bank energies alongside with the firs...

The availability of open-source software is playing a remarkable role in the popularization of speech recognition and deep learning. Kaldi, for instance, is nowadays an established framework used to develop state-of-the-art speech recognizers. PyTorch is used to build neural networks with the Python language and has recently spawn tremendous intere...

The availability of open-source software is playing a remarkable role in the popularization of speech recognition and deep learning. Kaldi, for instance, is nowadays an established framework used to develop state-of-the-art speech recognizers. PyTorch is used to build neural networks with the Python language and has recently spawn tremendous intere...

Convolutional neural networks (CNN) have recently achieved state-of-the-art results in various applications. In the case of image recognition, an ideal model has to learn independently of the training data, both local dependencies between the three components (R,G,B) of a pixel, and the global relations describing edges or shapes, making it efficie...

Recurrent neural networks (RNN) are at the core of modern automatic speech recognition (ASR) systems. In particular, long-short term memory (LSTM) recurrent neu-ral networks have achieved state-of-the-art results in many speech recognition tasks, due to their efficient representation of long and short term dependencies in sequences of interdependen...

Recently, the connectionist temporal classification (CTC) model coupled with recurrent (RNN) or convolutional neural networks (CNN), made it easier to train speech recognition systems in an end-to-end fashion. However in real-valued models, time frame components such as mel-filter-bank energies and the cepstral coefficients obtained from them, toge...

Recurrent neural networks (RNNs) are powerful architectures to model sequential data, due to their capability to learn short and long-term dependencies between the basic elements of a sequence. Nonetheless, popular tasks such as speech or images recognition, involve multi-dimensional input features that are characterized by strong internal dependen...

Recently, the connectionist temporal classification (CTC) model coupled with recurrent (RNN) or convolutional neural networks (CNN), made it easier to train speech recognition systems in an end-to-end fashion. However in real-valued models , time frame components such as mel-filter-bank energies and the cepstral coefficients obtained from them, tog...

Deep Neural Networks (DNN) received a great interest from researchers due to their capability to construct robust abstract representations of heterogeneous documents in a latent subspace. Nonetheless, mere real-valued deep neural networks require an appropriate adaptation, such as the con-volution process, to capture latent relations between input...

In the last decades, encoder-decoders or autoen-coders (AE) have received a great interest from researchers due to their capability to construct robust representations of documents in a low dimensional sub-space. Nonetheless, autoencoders reveal little in way of spoken document internal structure by only considering words or topics contained in the...