Tiago H. Falk

Heitor R. Guimarães, Arthur Pimentel, Anderson R. Avila et al.

Self-supervised speech representation learning aims to extract meaningful factors from the speech signal that can later be used across different downstream tasks, such as speech and/or emotion recognition. Existing models, such as HuBERT, however, can be fairly large thus may not be suitable for edge speech applications. Moreover, realistic applications typically involve speech corrupted by noise and room reverberation, hence models need to provide representations that are robust to such environmental factors. In this study, we build on the so-called DistilHuBERT model, which distils HuBERT to a fraction of its original size, with three modifications, namely: (i) augment the training data with noise and reverberation, while the student model needs to distill the clean representations from the teacher model; (ii) introduce a curriculum learning approach where increasing levels of noise are introduced as the model trains, thus helping with convergence and with the creation of more robust representations; and (iii) introduce a multi-task learning approach where the model also reconstructs the clean waveform jointly with the distillation task, thus also acting as an enhancement step to ensure additional environment robustness to the representation. Experiments on three SUPERB tasks show the advantages of the proposed method not only relative to the original DistilHuBERT, but also to the original HuBERT, thus showing the advantages of the proposed method for ``in the wild'' edge speech applications.

Yi Zhu, Mohamed Imoussaïne-Aïkous, Carolyn Côté-Lussier et al.

With advances seen in deep learning, voice-based applications are burgeoning, ranging from personal assistants, affective computing, to remote disease diagnostics. As the voice contains both linguistic and para-linguistic information (e.g., vocal pitch, intonation, speech rate, loudness), there is growing interest in voice anonymization to preserve speaker privacy and identity. Voice privacy challenges have emerged over the last few years and focus has been placed on removing speaker identity while keeping linguistic content intact. For affective computing and disease monitoring applications, however, the para-linguistic content may be more critical. Unfortunately, the effects that anonymization may have on these systems are still largely unknown. In this paper, we fill this gap and focus on one particular health monitoring application: speech-based COVID-19 diagnosis. We test three anonymization methods and their impact on five different state-of-the-art COVID-19 diagnostic systems using three public datasets. We validate the effectiveness of the anonymization methods, compare their computational complexity, and quantify the impact across different testing scenarios for both within- and across-dataset conditions. Additionally, we provided a comprehensive evaluation of the importance of different speech aspects for diagnostics and showed how they are affected by different types of anonymizers. Lastly, we show the benefits of using anonymized external data as a data augmentation tool to help recover some of the COVID-19 diagnostic accuracy loss seen with anonymization.

Yi Zhu, Saurabh Powar, Tiago H. Falk

Existing deepfake speech detection systems lack generalizability to unseen attacks (i.e., samples generated by generative algorithms not seen during training). Recent studies have explored the use of universal speech representations to tackle this issue and have obtained inspiring results. These works, however, have focused on innovating downstream classifiers while leaving the representation itself untouched. In this study, we argue that characterizing the long-term temporal dynamics of these representations is crucial for generalizability and propose a new method to assess representation dynamics. Indeed, we show that different generative models generate similar representation dynamics patterns with our proposed method. Experiments on the ASVspoof 2019 and 2021 datasets validate the benefits of the proposed method to detect deepfakes from methods unseen during training, significantly improving on several benchmark methods.

Yi Zhu, Tiago Falk

Speech is known to carry health-related attributes, which has emerged as a novel venue for remote and long-term health monitoring. However, existing models are usually tailored for a specific type of disease, and have been shown to lack generalizability across datasets. Furthermore, concerns have been raised recently towards the leakage of speaker identity from health embeddings. To mitigate these limitations, we propose WavRx, a speech health diagnostics model that captures the respiration and articulation related dynamics from a universal speech representation. Our in-domain and cross-domain experiments on six pathological speech datasets demonstrate WavRx as a new state-of-the-art health diagnostic model. Furthermore, we show that the amount of speaker identity entailed in the WavRx health embeddings is significantly reduced without extra guidance during training. An in-depth analysis of the model was performed, thus providing physiological interpretation of its improved generalizability and privacy-preserving ability.

Vamsikrishna Chemudupati, Marzieh Tahaei, Heitor Guimaraes et al.

Large self-supervised pre-trained speech models have achieved remarkable success across various speech-processing tasks. The self-supervised training of these models leads to universal speech representations that can be used for different downstream tasks, ranging from automatic speech recognition (ASR) to speaker identification. Recently, Whisper, a transformer-based model was proposed and trained on large amount of weakly supervised data for ASR; it outperformed several state-of-the-art self-supervised models. Given the superiority of Whisper for ASR, in this paper we explore the transferability of the representation for four other speech tasks in SUPERB benchmark. Moreover, we explore the robustness of Whisper representation for ``in the wild'' tasks where speech is corrupted by environment noise and room reverberation. Experimental results show Whisper achieves promising results across tasks and environmental conditions, thus showing potential for cross-task real-world deployment.

Andrew Perkis, Christian Timmerer, Sabina Baraković et al.

With the coming of age of virtual/augmented reality and interactive media, numerous definitions, frameworks, and models of immersion have emerged across different fields ranging from computer graphics to literary works. Immersion is oftentimes used interchangeably with presence as both concepts are closely related. However, there are noticeable interdisciplinary differences regarding definitions, scope, and constituents that are required to be addressed so that a coherent understanding of the concepts can be achieved. Such consensus is vital for paving the directionality of the future of immersive media experiences (IMEx) and all related matters. The aim of this white paper is to provide a survey of definitions of immersion and presence which leads to a definition of immersive media experience (IMEx). The Quality of Experience (QoE) for immersive media is described by establishing a relationship between the concepts of QoE and IMEx followed by application areas of immersive media experience. Influencing factors on immersive media experience are elaborated as well as the assessment of immersive media experience. Finally, standardization activities related to IMEx are highlighted and the white paper is concluded with an outlook related to future developments.

Karel Mundnich, Brandon M. Booth, Michelle L'Hommedieu et al.

We present a novel longitudinal multimodal corpus of physiological and behavioral data collected from direct clinical providers in a hospital workplace. We designed the study to investigate the use of off-the-shelf wearable and environmental sensors to understand individual-specific constructs such as job performance, interpersonal interaction, and well-being of hospital workers over time in their natural day-to-day job settings. We collected behavioral and physiological data from $n = 212$ participants through Internet-of-Things Bluetooth data hubs, wearable sensors (including a wristband, a biometrics-tracking garment, a smartphone, and an audio-feature recorder), together with a battery of surveys to assess personality traits, behavioral states, job performance, and well-being over time. Besides the default use of the data set, we envision several novel research opportunities and potential applications, including multi-modal and multi-task behavioral modeling, authentication through biometrics, and privacy-aware and privacy-preserving machine learning.

Joao Monteiro, Isabela Albuquerque, Jahangir Alam et al.

In this contribution, we augment the metric learning setting by introducing a parametric pseudo-distance, trained jointly with the encoder. Several interpretations are thus drawn for the learned distance-like model's output. We first show it approximates a likelihood ratio which can be used for hypothesis tests, and that it further induces a large divergence across the joint distributions of pairs of examples from the same and from different classes. Evaluation is performed under the verification setting consisting of determining whether sets of examples belong to the same class, even if such classes are novel and were never presented to the model during training. Empirical evaluation shows such method defines an end-to-end approach for the verification problem, able to attain better performance than simple scorers such as those based on cosine similarity and further outperforming widely used downstream classifiers. We further observe training is much simplified under the proposed approach compared to metric learning with actual distances, requiring no complex scheme to harvest pairs of examples.

Isabela Albuquerque, João Monteiro, Mohammad Darvishi et al.

Supervised learning results typically rely on assumptions of i.i.d. data. Unfortunately, those assumptions are commonly violated in practice. In this work, we tackle such problem by focusing on domain generalization: a formalization where the data generating process at test time may yield samples from never-before-seen domains (distributions). Our work relies on the following lemma: by minimizing a notion of discrepancy between all pairs from a set of given domains, we also minimize the discrepancy between any pairs of mixtures of domains. Using this result, we derive a generalization bound for our setting. We then show that low risk over unseen domains can be achieved by representing the data in a space where (i) the training distributions are indistinguishable, and (ii) relevant information for the task at hand is preserved. Minimizing the terms in our bound yields an adversarial formulation which estimates and minimizes pairwise discrepancies. We validate our proposed strategy on standard domain generalization benchmarks, outperforming a number of recently introduced methods. Notably, we tackle a real-world application where the underlying data corresponds to multi-channel electroencephalography time series from different subjects, each considered as a distinct domain.

Isabela Albuquerque, João Monteiro, Olivier Rosanne et al.

Assessment of mental workload in real-world conditions is key to ensure the performance of workers executing tasks that demand sustained attention. Previous literature has employed electroencephalography (EEG) to this end despite having observed that EEG correlates of mental workload vary across subjects and physical strain, thus making it difficult to devise models capable of simultaneously presenting reliable performance across users. Domain adaptation consists of a set of strategies that aim at allowing for improving machine learning systems performance on unseen data at training time. Such methods, however, might rely on assumptions over the considered data distributions, which typically do not hold for applications of EEG data. Motivated by this observation, in this work we propose a strategy to estimate two types of discrepancies between multiple data distributions, namely marginal and conditional shifts, observed on data collected from different subjects. Besides shedding light on the assumptions that hold for a particular dataset, the estimates of statistical shifts obtained with the proposed approach can be used for investigating other aspects of a machine learning pipeline, such as quantitatively assessing the effectiveness of domain adaptation strategies. In particular, we consider EEG data collected from individuals performing mental tasks while running on a treadmill and pedaling on a stationary bike and explore the effects of different normalization strategies commonly used to mitigate cross-subject variability. We show the effects that different normalization schemes have on statistical shifts and their relationship with the accuracy of mental workload prediction as assessed on unseen participants at training time.

Isabela Albuquerque, João Monteiro, Thang Doan et al.

Recent literature has demonstrated promising results for training Generative Adversarial Networks by employing a set of discriminators, in contrast to the traditional game involving one generator against a single adversary. Such methods perform single-objective optimization on some simple consolidation of the losses, e.g. an arithmetic average. In this work, we revisit the multiple-discriminator setting by framing the simultaneous minimization of losses provided by different models as a multi-objective optimization problem. Specifically, we evaluate the performance of multiple gradient descent and the hypervolume maximization algorithm on a number of different datasets. Moreover, we argue that the previously proposed methods and hypervolume maximization can all be seen as variations of multiple gradient descent in which the update direction can be computed efficiently. Our results indicate that hypervolume maximization presents a better compromise between sample quality and computational cost than previous methods.

Isabela Albuquerque, João Monteiro, Tiago H. Falk

In this work, we introduce a two-step framework for generative modeling of temporal data. Specifically, the generative adversarial networks (GANs) setting is employed to generate synthetic scenes of moving objects. To do so, we propose a two-step training scheme within which: a generator of static frames is trained first. Afterwards, a recurrent model is trained with the goal of providing a sequence of inputs to the previously trained frames generator, thus yielding scenes which look natural. The adversarial setting is employed in both training steps. However, with the aim of avoiding known training instabilities in GANs, a multiple discriminator approach is used to train both models. Results in the studied video dataset indicate that, by employing such an approach, the recurrent part is able to learn how to coherently navigate the image manifold induced by the frames generator, thus yielding more natural-looking scenes.

Yannick Roy, Hubert Banville, Isabela Albuquerque et al.

Electroencephalography (EEG) is a complex signal and can require several years of training to be correctly interpreted. Recently, deep learning (DL) has shown great promise in helping make sense of EEG signals due to its capacity to learn good feature representations from raw data. Whether DL truly presents advantages as compared to more traditional EEG processing approaches, however, remains an open question. In this work, we review 156 papers that apply DL to EEG, published between January 2010 and July 2018, and spanning different application domains such as epilepsy, sleep, brain-computer interfacing, and cognitive and affective monitoring. We extract trends and highlight interesting approaches in order to inform future research and formulate recommendations. Various data items were extracted for each study pertaining to 1) the data, 2) the preprocessing methodology, 3) the DL design choices, 4) the results, and 5) the reproducibility of the experiments. Our analysis reveals that the amount of EEG data used across studies varies from less than ten minutes to thousands of hours. As for the model, 40% of the studies used convolutional neural networks (CNNs), while 14% used recurrent neural networks (RNNs), most often with a total of 3 to 10 layers. Moreover, almost one-half of the studies trained their models on raw or preprocessed EEG time series. Finally, the median gain in accuracy of DL approaches over traditional baselines was 5.4% across all relevant studies. More importantly, however, we noticed studies often suffer from poor reproducibility: a majority of papers would be hard or impossible to reproduce given the unavailability of their data and code. To help the field progress, we provide a list of recommendations for future studies and we make our summary table of DL and EEG papers available and invite the community to contribute.

João Monteiro, Isabela Albuquerque, Zahid Akhtar et al.

Modern applications of artificial neural networks have yielded remarkable performance gains in a wide range of tasks. However, recent studies have discovered that such modelling strategy is vulnerable to Adversarial Examples, i.e. examples with subtle perturbations often too small and imperceptible to humans, but that can easily fool neural networks. Defense techniques against adversarial examples have been proposed, but ensuring robust performance against varying or novel types of attacks remains an open problem. In this work, we focus on the detection setting, in which case attackers become identifiable while models remain vulnerable. Particularly, we employ the decision layer of independently trained models as features for posterior detection. The proposed framework does not require any prior knowledge of adversarial examples generation techniques, and can be directly employed along with unmodified off-the-shelf models. Experiments on the standard MNIST and CIFAR10 datasets deliver empirical evidence that such detection approach generalizes well across not only different adversarial examples generation methods but also quality degradation attacks. Non-linear binary classifiers trained on top of our proposed features can achieve a high detection rate (>90%) in a set of white-box attacks and maintain such performance when tested against unseen attacks.

Joao Felipe Santos, Tiago H. Falk

In this paper, we propose a model to perform speech dereverberation by estimating its spectral magnitude from the reverberant counterpart. Our models are capable of extracting features that take into account both short and long-term dependencies in the signal through a convolutional encoder (which extracts features from a short, bounded context of frames) and a recurrent neural network for extracting long-term information. Our model outperforms a recently proposed model that uses different context information depending on the reverberation time, without requiring any sort of additional input, yielding improvements of up to 0.4 on PESQ, 0.3 on STOI, and 1.0 on POLQA relative to reverberant speech. We also show our model is able to generalize to real room impulse responses even when only trained with simulated room impulse responses, different speakers, and high reverberation times. Lastly, listening tests show the proposed method outperforming benchmark models in reduction of perceived reverberation.

M. Senoussaoui, J. F. Santos, T. H. Falk

Reverberation, especially in large rooms, severely degrades speech recognition performance and speech intelligibility. Since direct measurement of room characteristics is usually not possible, blind estimation of reverberation-related metrics such as the reverberation time (RT) and the direct-to-reverberant energy ratio (DRR) can be valuable information to speech recognition and enhancement algorithms operating in enclosed environments. The objective of this work is to evaluate the performance of five variants of blind RT and DRR estimators based on a modulation spectrum representation of reverberant speech with single- and multi-channel speech data. These models are all based on variants of the so-called Speech-to-Reverberation Modulation Energy Ratio (SRMR). We show that these measures outperform a state-of-the-art baseline based on maximum-likelihood estimation of sound decay rates in terms of root-mean square error (RMSE), as well as Pearson correlation. Compared to the baseline, the best proposed measure, called NSRMR_k , achieves a 23% relative improvement in terms of RMSE and allows for relative correlation improvements ranging from 13% to 47% for RT prediction.