Search

En los últimos años se ha extendido el uso de técnicas de Inteligencia Artificial para la detección y diagnóstico temprano de enfermedades dado que el tratamiento preventivo o en etapas iniciales es mucho más efectivo y su coste mucho menor. Sin embargo no hay demasiados trabajos que aborden el uso de estas técnicas para estimar el nivel de dolor que sufren los pacientes de determinadas enfermedades. En este proyecto se plantea diseñar un sistema basado en técnicas de inteligencia artificial con datos multimodales para estimar el nivel de dolor y poder ajustar en su caso los tratamientos para eliminarlo o mitigarlo.

The destructive power of a landslide can seriously affect human beings and infrastructures. The prediction of this phenomenon is of great interest; however, it is a complex task in which traditional methods have limitations. In recent years, Artificial Intelligence has emerged as a successful alternative in the geological field. Most of the related works use classical machine learning algorithms to correlate the variables of the phenomenon and its occurrence. This requires large quantitative landslide datasets, collected and labeled manually, which is costly in terms of time and effort. In this work, we create an image dataset using an official landslide inventory, which we verified and updated based on journalistic information and interpretation of satellite images of the study area. The images cover the landslide crowns and the actual triggering values of the conditioning factors at the detail level (5 × 5 pixels). Our approach focuses on the specific location where the landslide starts and its proximity, unlike other works that consider the entire landslide area as the occurrence of the phenomenon. These images correspond to geological, geomorphological, hydrological and anthropological variables, which are stacked in a similar way to the channels of a conventional image to feed and train a convolutional neural network. Therefore, we improve the quality of the data and the representation of the phenomenon to obtain a more robust, reliable and accurate prediction model. The results indicate an average accuracy of 97.48%, which allows the generation of a landslide susceptibility map on the Aloag-Santo Domingo highway in Ecuador. This tool is useful for risk prevention and management in this area where small, medium and large landslides occur frequently.

Anomaly detection is the identification of events or observations that deviate from the expected behaviour of a given set of data. Its main application is the prediction of possible technical failures. In particular, anomaly detection on supercomputers is a difficult problem to solve due to the large scale of the systems and the large number of components. Most research works in this field employ machine learning methods and regression models in a supervised fashion, which implies the need for a large amount of labelled data to train such systems. This work proposes the use of autoencoder models, allowing the problem to be approached with semi-supervised learning techniques. Two different model training approaches are compared. The former is a model trained with data from all the nodes of a supercomputer. In the latter approach, observing significant differences between nodes, one model is trained for each node. The results are analysed by evaluating the positive and negative aspects of each approach. On the other hand, a replica of the Marconi 100 supercomputer is developed in a virtual reality environment that allows the data from each node to be visualised at the same time.

Pain assessment is a critical aspect of healthcare, influencing timely interventions and patient well-being. Traditional pain evaluation methods often rely on subjective patient reports, leading to inaccuracies and disparities in treatment, especially for patients who present difficulties to communicate due to cognitive impairments. Our contributions are three-fold. Firstly, we analyze the correlations of the data extracted from biomedical sensors. Then, we use state-of-the-art computer vision techniques to analyze videos focusing on the facial expressions of the patients, both per-frame and using the temporal context. We compare them and provide a baseline for pain assessment methods using two popular benchmarks: UNBC-McMaster Shoulder Pain Expression Archive Database and BioVid Heat Pain Database. We achieved an accuracy of over 96% and over 94% for the F1 Score, recall and precision metrics in pain estimation using single frames with the UNBC-McMaster dataset, employing state-of-the-art computer vision techniques such as Transformer-based architectures for vision tasks. In addition, from the conclusions drawn from the study, future lines of work in this area are discussed.

Recent advances in facial recognition technology have achieved outstanding performance, but unconstrained face recognition remains an ongoing issue. Facial-image-quality-evaluation algorithms evaluate the quality of the input samples, providing crucial information about the accuracy of recognition decisions. By doing so, this can lead to improved results in challenging scenarios. In recent years, significant progress has been made in assessing the quality of facial images. The computation of quality scores has become highly precise and closely correlated with the model results. In this paper, we reviewed and analyzed the existing biases of cutting-edge quality-estimation techniques for face recognition. Our experimentation focused on the quality estimators developed by MagFace, FaceQNet, and SER-FIQ and were evaluated on the CelebA reference dataset. A study of bias in the face-recognition model was conducted by analyzing the quality scores presented in each article. This allowed for an examination of existing biases within both the quality estimators and the face-recognition models.

La gestión y seguimiento del desarrollo de proyectos o trabajos fin de grado TFG requiere de una gran cantidad de tiempo y suele realizarse de manera individual sin colaboración ni retroalimentación entre alumnos. Debido al desarrollo y presentación individual se ha perdido el espíritu colaborativo y de trabajo en equipo en una asignatura tan relevante. Los objetivos concretos del trabajo son: 1. Diseñar una metodología colaborativa y ágil para el seguimiento de los TFG, 2. Proponer y evaluar herramientas colaborativas en tareas concretas del TFG y 3. Comparar la nueva metodología con metodologías convencionales. Se utilizarán cuestionarios a los alumnos tutelados por miembros de la red como instrumentos para la recogida de información. En la fase de diseño de la experiencia se plantearán las herramientas y metodología a seguir con el grupo de alumnos tutelados. En concreto se plantea un enfoque innovador en la gestión de los estudiantes de TFG mediante el uso de herramientas colaborativas. Se van a diseñar una serie de materiales y recursos disponibles para los estudiantes.

In this project, we propose an action estimation pipeline based on the simultaneous recognition of the hands and the objects in the scene from an egocentric perspective. To this end, research on the state of the art in this field has been conducted. From the latest approaches, we have come to the conclusion that the hands are a key element from this point of view. An action consists of the interactions of the hands with the different objects in the scene therefore, the 2D positions of hands and objects are used to compute the object which is more likely to be used. The architecture used for achieving this goal is YOLO, as its prediction speed allows us to predict the actions fluently with good accuracy on the detected objects and hands. After reviewing the available datasets and generators for hand and object detection, different experiments have been conducted. The best results determined by PascalVOC metric have been used in the proposed pipeline.