Existing resources for Automatic Speech Recognition in Portuguese are mostly focused on Brazilian Portuguese, leaving European Portuguese (EP) and other varieties underexplored. To bridge this gap, we introduce CAMÕES, the first open framework for EP and other Portuguese varieties. It consists of (1) a comprehensive evaluation benchmark, including 46 h of EP test data spanning multiple domains; and (2) a collection of state-of-the-art models. For the latter, we consider multiple foundation models, evaluating their zero-shot and fine-tuned performances, as well as E-Branchformer models trained from scratch. A curated set of $\mathbf{4 2 5 h}$ of EP was used for both fine-tuning and training. Our results show comparable performance for EP between fine-tuned foundation models and the E-Branchformer. Furthermore, the best-performing models achieve relative improvements above 35% WER, compared to the strongest zero-shot foundation model, establishing a new state-of-the-art for EP and other varieties.

In the last months, there has been an increasing interest in developing reliable, cost-effective, immediate and easy to use machine learning based tools that can help health care operators, institutions, companies, etc. to optimize their screening campaigns.In this line, several initiatives emerged aimed at the automatic detection of COVID-19 from speech, breathing and coughs, with inconclusive preliminary results.The ComParE 2021 COVID-19 Cough Sub-challenge provides researchers from all over the world a suitable test-bed for the evaluation and comparison of their work.In this paper, we present the INESC-ID contribution to the ComParE 2021 COVID-19 Cough Sub-challenge.We leverage transfer learning to develop a set of three expert classifiers based on deep cough representation extractors.A calibrated decision-level fusion system provides the final classification of coughs recordings as either COVID-19 positive or negative.Results show unweighted average recalls of 72.3% and 69.3% in the development and test sets, respectively.Overall, the experimental assessment shows the potential of this approach although much more research on extended respiratory sounds datasets is needed.

Parkinson's disease (PD) is a progressive degenerative disorder of the central nervous system characterized by motor and non-motor symptoms. As the disease progresses, patients alternate periods in which motor symptoms are mitigated due to medication intake (ON state) and periods with motor complications (OFF state). The time that patients spend in the OFF condition is currently the main parameter employed to assess pharmacological interventions and to evaluate the efficacy of different active principles. In this work, we present a system that combines automatic speech processing and deep learning techniques to classify the medication state of PD patients by leveraging personal speech-based bio-markers. We devise a speaker-dependent approach and investigate the relevance of different acoustic-prosodic feature sets. Results show an accuracy of 90.54% in a test task with mixed speech and an accuracy of 95.27% in a semi-spontaneous speech task. Overall, the experimental assessment shows the potentials of this approach towards the development of reliable, remote daily monitoring and scheduling of medication intake of PD patients.

Automatic personality analysis has gained attention in the last years as a fundamental dimension in human-To-human and human-To-machine interaction. However, it still suffers from limited number and size of speech corpora for specific domains, such as the assessment of children's personality. This paper investigates a semi-supervised training approach to tackle this scenario. We devise an experimental setup with age and language mismatch and two training sets: A small labeled training set from the Interspeech 2012 Personality Sub-challenge, containing French adult speech labeled with personality OCEAN traits, and a large unlabeled training set of Portuguese children's speech. As test set, a corpus of Portuguese children's speech labeled with OCEAN traits is used. Based on this setting, we investigate a weak supervision approach that iteratively refines an initial model trained with the labeled data-set using the unlabeled data-set. We also investigate knowledge-based features, which leverage expert knowledge in acoustic-prosodic cues and thus need no extra data. Results show that, despite the large mismatch imposed by language and age differences, it is possible to attain improvements with these techniques, pointing both to the benefits of using a weak supervision and expert-based acoustic-prosodic features across age and language.

In the last years, support vector machines (SVMs) have shown excellent performance in many applications, especially in the presence of noise. In particular, SVMs offer several advantages over artificial neural networks (ANNs) that have attracted the attention of the speech processing community. Nevertheless, their high computational requirements prevent them from being used in practice in automatic speech recognition (ASR), where ANNs have proven to be successful. The high complexity of SVMs in this context arises from the use of huge speech training databases with millions of samples and highly overlapped classes. This paper suggests the use of a weighted least squares (WLS) training procedure that facilitates the possibility of imposing a compact semiparametric model on the SVM, which results in a dramatic complexity reduction. Such a complexity reduction with respect to conventional SVMs, which is between two and three orders of magnitude, allows the proposed hybrid WLS-SVC/HMM system to perform real-time speech decoding on a connected-digit recognition task (SpeechDat Spanish database). The experimental evaluation of the proposed system shows encouraging performance levels in clean and noisy conditions, although further improvements are required to reach the maturity level of current context-dependent HMM-based recognizers.

Hybrid speech recognizers, where the estimation of the emission pdf of the states of hidden Markov models (HMMs), usually carried out using Gaussian mixture models (GMMs), is substituted by artificial neural networks (ANNs) have several advantages over the classical systems. However, to obtain performance improvements, the computational requirements are heavily increased because of the need to train the ANN. Departing from the observation of the remarkable skewness of speech data, this paper proposes sifting out the training set and balancing the amount of samples per class. With this method, the training time has been reduced 18 times while obtaining performances similar to or even better than those with the whole database, especially in noisy environments. However, the application of these reduced sets is not straightforward. To avoid the mismatch between training and testing conditions created by the modification of the distribution of the training data, a proper scaling of the a posteriori probabilities obtained and a resizing of the context window need to be performed as demonstrated in this paper.