Search

Speech Emotion Recognition (SER) is crucial for enabling computers to understand the emotions conveyed in human communication. With recent advancements in Deep Learning (DL), the performance of SER models has significantly improved. However, designing an optimal DL architecture requires specialised knowledge and experimental assessments. Fortunately, Neural Architecture Search (NAS) provides a potential solution for automatically determining the best DL model. The Differentiable Architecture Search (DARTS) is a particularly efficient method for discovering optimal models. This study presents emoDARTS, a DARTS-optimised joint CNN and Sequential Neural Network (SeqNN: LSTM, RNN) architecture that enhances SER performance. The literature supports the selection of CNN and LSTM coupling to improve performance. While DARTS has previously been used to choose CNN and LSTM operations independently, our technique adds a novel mechanism for selecting CNN and SeqNN operations in conjunction using DARTS. Unlike earlier work, we do not impose limits on the layer order of the CNN. Instead, we let DARTS choose the best layer order inside the DARTS cell. We demonstrate that emoDARTS outperforms conventionally designed CNN-LSTM models and surpasses the best-reported SER results achieved through DARTS on CNN-LSTM by evaluating our approach on the IEMOCAP, MSP-IMPROV, and MSP-Podcast datasets.

In autonomous, as well as manually operated vehicles, monitoring the driver visual attention provides useful information about the behavior, intent and vigilance level of the driver. The gaze of the driver can be formulated in terms of a probabilistic visual map representing the region around which the driver’s attention is focused. The area of the estimated region changes based on the level of confidence of the estimation. This paper proposes a framework based on convolutional neural networks (CNNs) that takes the head pose and the eye appearance of the driver as inputs, and creates a fusion model that estimates the driver’s gaze on a 2D grid. The model contains upsampling layers to create estimations at multiple resolutions. The model is trained using data collected from 59 subjects with continuous recordings where the subject looks at a moving target in a parked car, and glances at a set of markers inside the car while driving the vehicle and while the car is parked. Our fusion framework provides superior performance than unimodal systems trained exclusively with head pose or eye appearance information. It estimates the gaze region with the target location lying within the 75% confidence region with an accuracy of 92.54%.

Baccharis ulicina es una maleza ampliamente distribuida en los pastizales de la zona semiárida argentina. A fin de evaluar sus posibles efectos alelopáticos, se utilizaron extractos acuosos de B. ulicina (hoja, tallo, raíz y planta entera) a dos concentraciones (50 y 150 g tejido/L agua) sobre la germinación y crecimiento inicial de Avena sativa, Lolium perenne y Raphanus sativus. Para cada especie se regaron 100 semillas y se mantuvieron a 25°C y oscuridad. Se registraron la germinación de las semillas y la longitud de coleóptilo / hipocótile (C/H) y radícula (R). Con los datos obtenidos se determinó: porcentaje de germinación, coeficiente de velocidad de germinación, vigor de la germinación, inhibición de la germinación, inhibición de la longitud de la radícula y del coleóptilo / hipocótile e índice de vigor de radícula y de coleóptilo / hipocótile. Los extractos con alta concentración redujeron la germinación en promedio 65% mientras que inhibieron la longitud en 58,8% (C/H) y 93,3% (R) para A. sativa; 71,5% (C/H) y 87,9% (R) para L. perenne, y 93,8% (C/H) y 94,5% (R) para R. sativus. Se registró un evidente efecto fitotóxico de B. ulicina sobre las especies evaluadas, indicando que la alelopatía sería una estrategia ecológica de la especie.

Studying the facial expressions of humans has been one of the major applications of computer vision. An open question is whether common machine learning techniques can also be used to track behaviors of animals, which is a less explored research problem. Since animals are not capable of verbal communication, computer vision solutions can provide valuable information to track the animal’s state. We are particularly interested in pain neurobiology research, where rodent models are extensively used to investigate pain interventions. A grimace scale is used to understand the suffering of a mouse in the presence of interventions, which is inferred from various facial features such as the shape of the eyes and ears. In this work, we automate the prediction of the grimace scale on white furred mice using a machine learning approach, following the same principles used for human facial expression recognition: face detection, landmark region extraction, and expression recognition. We demonstrate the use of the you only look once (YOLO) framework for face detection of the mice with outstanding results. For eye region extraction and grimace pain prediction, we propose a novel structure based on a dilated convolutional network. The experimental results are promising, showing that it is possible to differentiate among the pain scale of the mice.

This paper aims to create neutral reference models from synthetic speech to contrast the emotional content of a speech signal. Modeling emotional behaviors is a challenging task due to the variability in perceiving and describing emotions. Previous studies have indicated that relative assessments are more reliable than absolute assessments. These studies suggest that having a reference signal with known emotional content (e.g., neutral emotion) to compare a target sentence may produce more reliable metrics to identify emotional segments. Ideally, we would like to have an emotionally neutral sentence with the same lexical content as the target sentence where their contents are timely aligned. In this fictitious scenario, we would be able to identify localized emotional cues by contrasting frame-by-frame the acoustic features of the target and reference sentences. This paper explores the idea of building these reference sentences leveraging the advances in speech synthesis. This paper builds a synthetic speech signal that conveys the same lexical information and is timely aligned with the target sentence in the database. Since it is expected that a single synthetic speech will not capture the full range of variability observed in neutral speech, we build multiple synthetic sentences using various voices and text-to-speech approaches. This paper analyzes whether the synthesized signals provide valid template references to describe neutral speech using feature analysis and perceptual evaluation. Finally, we demonstrate how this framework can be used in emotion recognition, achieving improvements over classifiers trained with the state-of-the-art features in detecting low versus high levels of arousal and valence.

Wavelet Packets (WPs) bases are explored seeking new discriminative features for texture indexing. The task of WP feature design is formulated as a learning decision problem by selecting the filter-bank structure of a basis (within a WPs family) that offers an optimal balance between estimation and approximation errors. To address this problem, a computationally efficient algorithm is adopted that uses the tree-structure of the WPs collection and the Kullback-Leibler divergence as a discrimination criterion. The adaptive nature of the proposed solution is demonstrated in synthetic and real data scenarios. With synthetic data, we demonstrate that the proposed features can identify discriminative bands, which is not possible with standard wavelet decomposition. With data with real textures, we show performance improvements with respect to the conventional Wavelet-based decomposition used under the same conditions and model assumptions.

Introduction: Patients with type 2 diabetes (T2D) have an increased risk of developing cardiovascular complications, which is also the main cause of death in this population. Certain interventions like lifestyle changes and some pharmacological groups including statins, aspirin use in secondary prevention patients, and antidiabetic drugs such as SGLT2 inhibitors and GLP-1 receptor agonists (GLP-1 RA) have shown reduction in cardiovascular events for wich they are within the first line of recommendations in the main treatment guidelines. Objectives: Asses adherence to a healthy lifestyle and the use of pharmacological treatments with proven cardiovascular benefit in patients with T2D who are being followed by cardiologists. Study design and methods: We carried out an observational, cross-sectional and multicenter study. Patients with DM2 that assisted to the cardiology office from May to July 2019, were consecutively registered. From the total p with DM2, we included in the registry, the first two p of each day. Healthy lifestyle was define as follow: 1) Physical activity = physical activity > 150 minutes/week 2) Waist circumference = < 88 cm in women and < 102 in men 3) without obesity = BMI < 30 kg/m2 4) Intake of > 5 servings of fruit and vegetables 5) no smoking. Pharmacological treatments with cardiovascular benefit: 1) Statin use 2) SGLT2i and GLP-1 RA use 3) Aspirin use in patients with cardiovascular disease.

Introduction: The treatment of chronic non-communicable diseases, including type 2 diabetes mellitus (DM2), generally requires changes in lifestyle and pharmacological measures. Lack of adherence is a frequent problem around de world, ranging between 30 to 50% in different studies. Non-adherent patients have a higher incidence of cardiovascular events and higher mortality compared to adherent patients. Therefore, it increases costs in the health systems. Objective: Analyze the adherence to pharmacological treatment indicated in a group of patients with DM2 and high cardiovascular risk. Assess the differences in clinical characteristics and scopes of therapeutic goals in the adherent population in comparison with non-adherent patients. Study design and methods: We carried out an observational, cross-sectional and multicenter study. Patients with DM2 that assisted to the cardiology office from May to July 2019, were consecutively registered. From the total p with DM2, we included in the registry, the first two p of each day. Adherence was assessed using the Morisky-Green test, which consists of 4 questions with two possible answers options: YES/NO. We defined adherence when the 4 answers were affirmative. We consider a non-adherent patient when presenting 1 or more negative answers. Therapeutic goals were defined: blood pressure under 140/90 mmHg, 2) HbA1c < 7% and 3) LDL cholesterol < 100 in patients without Cardiovascular disease (CVD) and < 70 mg/dl in patients with CVD. 4) triple goal: points 1,2 and 3 within the corresponding values.

Introduction: arterial hypertension (AH) and diabetes mellitus (DM) are chronic medical conditions of high prevalence, which are often associated. Conclusions: AH increases the progression and development of chronic micro and macro vascular complications related to DM. The impact of Hypertension treatment is highly significant in reducing morbidity and mortality in people with diabetes mellitus. Therefore, proper early treatment and therapeutic target goals should be individualized according to age group, comorbidities, and target organ damage. The goal is to achieve Blood Pressure (BP)

Acoustic metrics extracted from speech have the potential to serve as novel biomarkers for a variety of neurological and neurodevelopmental conditions, as is evidenced by the rapidly growing corpus of research articles studying the links between brain impairments and speech. In this paper, we discuss the advantages and the disadvantages of speech biomarkers and the various challenges in the design and the implementation of portable speech-based diagnostic and assessment tools. Furthermore, we provide a case study, presenting our experiences in developing an assessment tool for the detection of mild traumatic brain injuries (concussions) and discuss the challenges in obtaining and analyzing large sets of speech recordings that can be used to study the impact of brain injuries on vocal features.