Automatic classification of dog barking using deep learning.

Barking and other dog vocalizations have acoustic properties related to emotions, physiological reactions, attitudes, or some particular internal states. In the field of intelligent audio analysis, researchers use methods based on signal processing and machine learning to analyze the digitized acoustic signals' properties and obtain relevant information. The present work describes a method to classify the identity, breed, age, sex, and context associated with each bark. This information can support the decisions of people who regularly interact with animals, such as dog trainers, veterinarians, rescuers, police, people with visual impairment. Our approach uses deep neural networks to generate trained models for each classification task. We worked with 19,643 barks recorded from 113 dogs of different breeds, ages and sexes. Our methodology consists of three stages. First, the pre-processing stage prepares the data and transforms it into the appropriate format for each classification model. Second, the characterization stage evaluates different representation models to identify the most suitable for each task. Third, the classification stage trains each classification model and selects the best hyperparameters. After tuning and training each model, we evaluated its performance. We analyzed the most relevant features extracted from the audio and the most appropriate deep neural network architecture for that feature type. Even if the application of our method is not ready for being used in ethological practice, our evaluation showed an outstanding performance of the proposed method, surpassing previous research results on this topic, providing the basis for further technological development.

Machine Learning-Based Sensor Data Fusion for Animal Monitoring: Scoping Review.

The development of technology, such as the Internet of Things and artificial intelligence, has significantly advanced many fields of study. Animal research is no exception, as these technologies have enabled data collection through various sensing devices. Advanced computer systems equipped with artificial intelligence capabilities can process these data, allowing researchers to identify significant behaviors related to the detection of illnesses, discerning the emotional state of the animals, and even recognizing individual animal identities. This review includes articles in the English language published between 2011 and 2022. A total of 263 articles were retrieved, and after applying inclusion criteria, only 23 were deemed eligible for analysis. Sensor fusion algorithms were categorized into three levels: Raw or low (26%), Feature or medium (39%), and Decision or high (34%). Most articles focused on posture and activity detection, and the target species were primarily cows (32%) and horses (12%) in the three levels of fusion. The accelerometer was present at all levels. The findings indicate that the study of sensor fusion applied to animals is still in its early stages and has yet to be fully explored. There is an opportunity to research the use of sensor fusion for combining movement data with biometric sensors to develop animal welfare applications. Overall, the integration of sensor fusion and machine learning algorithms can provide a more in-depth understanding of animal behavior and contribute to better animal welfare, production efficiency, and conservation efforts.

Digital Technology Supporting the Remote Human-Dog Interaction: Scoping Review.

For thousands of years, dogs have coexisted with humans and have been adopted as companion pets and working animals. The communication between humans and dogs has improved their coexistence and socialization; however, due to the nature of their activities, dogs and humans occasionally lose face-to-face contact. The purpose of this scoping review is to examine five essential aspects of current technology designed to support intentional communication between humans and dogs in scenarios where there is no face-to-face contact: (1) the technologies used, (2) the activity supported, (3) the interaction modality, (4) the evaluation procedures, and the results obtained, and (5) the main limitations. In addition, this article explores future directions for research and practice. The PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines were followed when conducting the review. Scopus (Elsevier), Springer-Link, IEEE Xplorer, ACM Digital Library, and Science Direct were used as data sources to retrieve information from January 2010 to March 2022. The titles and abstracts were individually reviewed by the authors (L.R.-V., I.E.E.-C., and H.P.-E.), and the full articles were then examined before a final inclusion determination. 15 (3%) out of the 571 records that were obtained met the requirements for inclusion. The most used technologies for dogs are: (1) 71% of technologies focused on generating messages are wearable devices equipped with sensors (bite, tug, or gesture), (2) 60% of technologies focused on receiving messages are wearable devices equipped with vibrotactile actuators, and (3) 100% of technologies focused on bidirectional communication are videochats. 67% of the works are oriented to support search and assistance tasks. 80% of the works developed technology for one-way communication. 53% of the technologies have a haptic dog interaction modality, that is, there is an object that the dog must wear or manipulate in a certain way. All of the reported evaluations were pilot studies with positive feasibility results. Remote human-dog interaction technology holds significant promise and potential; however, more research is required to assess their usability and efficacy and to incorporate new technological developments.

A Generic Deep Learning Based Cough Analysis System From Clinically Validated Samples for Point-of-Need Covid-19 Test and Severity Levels.

In an attempt to reduce the infection rate of the COrona VIrus Disease-19 (Covid-19) countries around the world have echoed the exigency for an economical, accessible, point-of-need diagnostic test to identify Covid-19 carriers so that they (individuals who test positive) can be advised to self isolate rather than the entire community. Availability of a quick turn-around time diagnostic test would essentially mean that life, in general, can return to normality-at-large. In this regards, studies concurrent in time with ours have investigated different respiratory sounds, including cough, to recognise potential Covid-19 carriers. However, these studies lack clinical control and rely on Internet users confirming their test results in a web questionnaire (crowdsourcing) thus rendering their analysis inadequate. We seek to evaluate the detection performance of a primary screening tool of Covid-19 solely based on the cough sound from 8,380 clinically validated samples with laboratory molecular-test (2,339 Covid-19 positive and 6,041 Covid-19 negative) under quantitative RT-PCR (qRT-PCR) from certified laboratories. All collected samples were clinically labelled, i.e., Covid-19 positive or negative, according to the results in addition to the disease severity based on the qRT-PCR threshold cycle (Ct) and lymphocytes count from the patients. Our proposed generic method is an algorithm based on Empirical Mode Decomposition (EMD) for cough sound detection with subsequent classification based on a tensor of audio sonographs and deep artificial neural network classifier with convolutional layers called 'DeepCough'. Two different versions of DeepCough based on the number of tensor dimensions, i.e., DeepCough2D and DeepCough3D, have been investigated. These methods have been deployed in a multi-platform prototype web-app 'CoughDetect'. Covid-19 recognition results rates achieved a promising AUC (Area Under Curve) of [Formula: see text] 98 . 80 % ± 0 . 83 % , sensitivity of [Formula: see text] 96 . 43 % ± 1 . 85 % , and specificity of [Formula: see text] 96 . 20 % ± 1 . 74 % and average AUC of [Formula: see text] 81 . 08 % ± 5 . 05 % for the recognition of three severity levels. Our proposed web tool as a point-of-need primary diagnostic test for Covid-19 facilitates the rapid detection of the infection. We believe it has the potential to significantly hamper the Covid-19 pandemic across the world.

HelperFriend, a Serious Game for Promoting Healthy Lifestyle Behaviors in Children: Design and Pilot Study.

BACKGROUND: The use of health games is a promising strategy for educating and promoting healthy lifestyle behaviors among children. OBJECTIVE: We aimed to describe the design and development of a serious game, called HelperFriend, and evaluate its feasibility, acceptability, and preliminary effects in children in a pilot study. HelperFriend is a vicarious experiential video game designed to promote 3 lifestyle behaviors among young children: physical activity, healthy eating, and socioemotional wellness. METHODS: Participants aged 8 to 11 years were recruited from an elementary school and randomized to receive a healthy lifestyle behavior educational talk (control) or play six 30-minute sessions with HelperFriend (intervention). Assessments were conducted at baseline (T0) and after the intervention (ie, 4 weeks) (T1). The primary outcome was gain in knowledge. The secondary outcomes were intention to conduct healthy behaviors, dietary intake, and player satisfaction. RESULTS: Knowledge scores of intervention group participants increased from T0 to T1 for physical activity (t14=2.01, P=.03), healthy eating (t14=3.14, P=.003), and socioemotional wellness (t14=2.75, P=.008). In addition, from T0 to T1, the intervention group improved their intention to perform physical activity (t14=2.82, P=.006), healthy eating (t14=3.44, P=.002), and socioemotional wellness (t14=2.65, P=.009); and there was a reduction in their intake of 13 unhealthy foods. HelperFriend was well received by intervention group. CONCLUSIONS: HelperFriend appears to be feasible and acceptable for young children. In addition, this game seems to be a viable tool to help improve the knowledge, the intention to conduct healthy behaviors, and the dietary intake of children; however, a well-powered randomized controlled trial is needed to prove the efficacy of HelperFriend.

Relationship Between Children's Enjoyment, User Experience Satisfaction, and Learning in a Serious Video Game for Nutrition Education: Empirical Pilot Study.

BACKGROUND: The design and use of serious video games for children have increased in recent years. To maximize the effects of these games, it is essential to understand the children's experiences through playing. Previous studies identified that enjoyment and user experience satisfaction of the players are principal factors that can influence the success of serious video games and the learning of their players. However, research about the relationship between enjoyment and user experience satisfaction with learning in children 8 to 10 years old is sparse. OBJECTIVE: We examined the relationship of enjoyment and user experience satisfaction with the learning of children aged 8 to 10 years while playing a serious video game for health, FoodRateMaster. This serious video game teaches children about the characteristics of healthy and unhealthy foods and how to identify them in their environment. METHODS: Children aged 8 to 10 years were recruited from a primary school in Mexico. Participants completed 12 individual gaming sessions with FoodRateMaster in 6 weeks. A food knowledge questionnaire was administered before and after game play to assess the players' food knowledge. In addition, after the gaming sessions, the children's enjoyment and user experience satisfaction were evaluated using the EGameFlow questionnaire and the Game User Experience Satisfaction Scale (GUESS) questionnaire. RESULTS: We found significant positive associations for children's (n=60) posttest knowledge with enjoyment (r58=0.36, P=.005) and user experience satisfaction (r58=0.27, P=.04). The children's posttest knowledge scores were also positively correlated with challenge (r58=0.38, P=.003), knowledge improvement (r58=0.38, P=.003), and goal clarity (r58=0.29, P=.02) EGameFlow subscales and with narrative (r58=0.35, P=.006), creative freedom (r58=0.26, P=.04), and visual esthetics (r58=0.32, P=.01) GUESS subscales. Regression analysis indicated that the EGameFlow (F7,52=2.74, P=.02, R2=0.27) and the GUESS (F8,51=2.20, P=.04, R2=0.26) ratings significantly predicted the children's posttest knowledge scores. EGameFlow challenge (ß=0.40, t52=2.17, P=.04) and knowledge improvement (ß=0.29, t52=2.06, P=.04) subscales significantly contributed to predicting children's learning. None of the GUESS subscales significantly contributed to predicting children's learning. CONCLUSIONS: The findings of this study suggest that both enjoyment and user experience satisfaction for children aged 8 to 10 years were positively correlated with their learning and that were significant predictors of it. Challenge, knowledge improvement, narrative, creative freedom, and visual esthetics subscales correlated positively with children's learning. In addition, challenge and knowledge improvement contributed to predicting their learning. These results are relevant to consider during the design stages of serious games developed for young children's learning purposes.