SDHAR-HOME: A Sensor Dataset for Human Activity Recognition at Home.

Nowadays, one of the most important objectives in health research is the improvement of the living conditions and well-being of the elderly, especially those who live alone. These people may experience undesired or dangerous situations in their daily life at home due to physical, sensorial or cognitive limitations, such as forgetting their medication or wrong eating habits. This work focuses on the development of a database in a home, through non-intrusive technology, where several users are residing by combining: a set of non-intrusive sensors which captures events that occur in the house, a positioning system through triangulation using beacons and a system for monitoring the user's state through activity wristbands. Two months of uninterrupted measurements were obtained on the daily habits of 2 people who live with a pet and receive sporadic visits, in which 18 different types of activities were labelled. In order to validate the data, a system for the real-time recognition of the activities carried out by these residents was developed using different current Deep Learning (DL) techniques based on neural networks, such as Recurrent Neural Networks (RNN), Long Short-Term Memory networks (LSTM) or Gated Recurrent Unit networks (GRU). A personalised prediction model was developed for each user, resulting in hit rates ranging from 88.29% to 90.91%. Finally, a data sharing algorithm has been developed to improve the generalisability of the model and to avoid overtraining the neural network.

Daily Human Activity Recognition Using Non-Intrusive Sensors.

In recent years, Artificial Intelligence Technologies (AIT) have been developed to improve the quality of life of the elderly and their safety in the home. This work focuses on developing a system capable of recognising the most usual activities in the daily life of an elderly person in real-time to enable a specialist to monitor the habits of this person, such as taking medication or eating the correct meals of the day. To this end, a prediction model has been developed based on recurrent neural networks, specifically on bidirectional LSTM networks, to obtain in real-time the activity being carried out by the individuals in their homes, based on the information provided by a set of different sensors installed at each person's home. The prediction model developed in this paper provides a 95.42% accuracy rate, improving the results of similar models currently in use. In order to obtain a reliable model with a high accuracy rate, a series of processing and filtering processes have been carried out on the data, such as a method based on a sliding window or a stacking and re-ordering algorithm, that are subsequently used to train the neural network, obtained from the public database CASAS.

Development and Usability Validation of a Social Robot Platform for Physical and Cognitive Stimulation in Elder Care Facilities.

This article shows our work for developing an elder care platform for social interaction and physical and cognitive stimulation using the Pepper robot and Android OS as clients, based on the knowledge acquired on our long-term social robotics research experience. The first results of the user's acceptance of the solution are presented in this article. The platform is able to provide different services to the user, such as information, news, games, exercises or music. The games, which have a bi-modal way of interacting (speech and a touch screen interface), have been designed for cognitive stimulation based on the items of the mini-mental state examination. The results of the user's performance are stored in a cloud database and can be reviewed by therapists through a web interface that also allows them to establish customized therapy plans for each user. The platform has been tested and validated, first using adult people and then deployed to an elder care facility where the robot has been interacting with users for a long period of time. The results and feedback received have shown that the robot can help to keep the users physically and mentally active as well as establish an emotional link between the user and the robot.

Gaze Control of a Robotic Head for Realistic Interaction With Humans.

When there is an interaction between a robot and a person, gaze control is very important for face-to-face communication. However, when a robot interacts with several people, neurorobotics plays an important role to determine the person to look at and those to pay attention to among the others. There are several factors which can influence the decision: who is speaking, who he/she is speaking to, where people are looking, if the user wants to attract attention, etc. This article presents a novel method to decide who to pay attention to when a robot interacts with several people. The proposed method is based on a competitive network that receives different stimuli (look, speak, pose, hoard conversation, habituation, etc.) that compete with each other to decide who to pay attention to. The dynamic nature of this neural network allows a smooth transition in the focus of attention to a significant change in stimuli. A conversation is created between different participants, replicating human behavior in the robot. The method deals with the problem of several interlocutors appearing and disappearing from the visual field of the robot. A robotic head has been designed and built and a virtual agent projected on the robot's face display has been integrated with the gaze control. Different experiments have been carried out with that robotic head integrated into a ROS architecture model. The work presents the analysis of the method, how the system has been integrated with the robotic head and the experiments and results obtained.

Integration of Computer Vision and Wireless Networks to Provide Indoor Positioning.

This work presents an integrated Indoor Positioning System which makes use of WiFi signals and RGB cameras, such as surveillance cameras, to track and identify people navigating in complex indoor environments. Previous works have often been based on WiFi, but accuracy is limited. Other works use computer vision, but the problem of identifying concrete persons relies on such techniques as face recognition, which are not useful if there are many unknown people, or where the robustness decreases when individuals are seen from different points of view. The solution presented in this paper is based on an accurate combination of smartphones along with RGB cameras, such as those used in surveillance infrastructures. WiFi signals from smartphones allow the persons present in the environment to be identified uniquely, while the data coming from the cameras allow the precision of location to be improved. The system is nonintrusive, and biometric data about subjects is not required. In this paper, the proposed method is fully described and experiments performed to test the system are detailed along with the results obtained.

Crack Detection in Concrete Tunnels Using a Gabor Filter Invariant to Rotation.

In this article, a system for the detection of cracks in concrete tunnel surfaces, based on image sensors, is presented. Both data acquisition and processing are covered. Linear cameras and proper lighting are used for data acquisition. The required resolution of the camera sensors and the number of cameras is discussed in terms of the crack size and the tunnel type. Data processing is done by applying a new method called Gabor filter invariant to rotation, allowing the detection of cracks in any direction. The parameter values of this filter are set by using a modified genetic algorithm based on the Differential Evolution optimization method. The detection of the pixels belonging to cracks is obtained to a balanced accuracy of 95.27%, thus improving the results of previous approaches.

Corrigendum: Virtual Avatar for Emotion Recognition in Patients with Schizophrenia: A Pilot Study.

[This corrects the article on p. 421 in vol. 10, PMID: 27616987.].

Virtual Avatar for Emotion Recognition in Patients with Schizophrenia: A Pilot Study.

Persons who suffer from schizophrenia have difficulties in recognizing emotions in others' facial expressions, which affects their capabilities for social interaction and hinders their social integration. Photographic images have traditionally been used to explore emotion recognition impairments in schizophrenia patients, but they lack of the dynamism that is inherent to facial expressiveness. In order to overcome those inconveniences, over the last years different authors have proposed the use of virtual avatars. In this work, we present the results of a pilot study that explored the possibilities of using a realistic-looking avatar for the assessment of emotion recognition deficits in patients who suffer from schizophrenia. In the study, 20 subjects with schizophrenia of long evolution and 20 control subjects were invited to recognize a set of facial expressions of emotions showed by both the said virtual avatar and static images. Our results show that schizophrenic patients exhibit recognition deficits in emotion recognition from facial expressions regardless the type of stimuli (avatar or images), and that those deficits are related with the psychopathology. Finally, some improvements in recognition rates (RRs) for the patient group when using the avatar were observed for sadness or surprise expressions, and they even outperform the control group in the recognition of the happiness expression. This leads to conclude that, apart from the dynamism of the shown expression, the RRs for schizophrenia patients when employing animated avatars may depend on other factors which need to be further explored.

Automated Tracking of Drosophila Specimens.

The fruit fly Drosophila Melanogaster has become a model organism in the study of neurobiology and behavior patterns. The analysis of the way the fly moves and its behavior is of great scientific interest for research on aspects such as drug tolerance, aggression or ageing in humans. In this article, a procedure for detecting, identifying and tracking numerous specimens of Drosophila by means of computer vision-based sensing systems is presented. This procedure allows dynamic information about each specimen to be collected at each moment, and then for its behavior to be quantitatively characterized. The proposed algorithm operates in three main steps: a pre-processing step, a detection and segmentation step, and tracking shape. The pre-processing and segmentation steps allow some limits of the image acquisition system and some visual artifacts (such as shadows and reflections) to be dealt with. The improvements introduced in the tracking step allow the problems corresponding to identity loss and swaps, caused by the interaction between individual flies, to be solved efficiently. Thus, a robust method that compares favorably to other existing methods is obtained.