In this paper, a novel dataset is introduced, designed to assess student attention within in-person classroom settings. This dataset encompasses RGB camera data, featuring multiple cameras per student to capture both posture and facial expressions, in addition to smartwatch sensor data for each individual. This dataset allows machine learning algorithms to be trained to predict attention and correlate it with emotion. A comprehensive suite of attention and emotion labels for each student is provided, generated through self-reporting as well as evaluations by four different experts. Our dataset uniquely combines facial and environmental camera data, smartwatch metrics, and includes underrepresented ethnicities in similar datasets, all within in-the-wild, in-person settings, making it the most comprehensive dataset of its kind currently available. The dataset presented offers an extensive and diverse collection of data pertaining to student interactions across different educational contexts, augmented with additional metadata from other tools. This initiative addresses existing deficiencies by offering a valuable resource for the analysis of student attention and emotion in face-to-face lessons.

A crucial issue in education is to properly define the assessment tools such that the learning process is improved. In this sense, several attempts can be found in the literature. Among them, competition has led to great results in the teaching-learning process. Although its use is somewhat controversial, when competition and cooperation are balanced, the competition encourages students to learn from their failures, share their knowledge to get better results, study in-depth the subject contents and put it in practice to solve the proposed challenge. This hypothesis has been proven with the results of our pilot study with students of the third course of the Degree in Robotics Engineering at the University of Alicante (Spain), within the subject of "Computer Vision", taken during the last academic course. As learning is the final goal, all the techniques presented in the subject can be used and no limits about the accuracy in the results are established. In this way, the students implement and test several techniques to improve the results. Moreover, the active implementation results in discussions about the suitability of the different techniques for the proposed challenge and the parameter values that improve their performance. As shown in the results, the quality of the proposed solutions presents a proper accuracy and the learning process has been improved as reflected in the students' marks.

In 2016 we proposed a new course syllabus for the laboratory sessions of the subject Artificial Vision and Robotics [1]. Our proposal included one activity where the students must implement an algorithm to autonomously drive a car in a simulated environment. To encourage the students to do their best, two measures were taken. First, the students must compete against each other upon the deadline, as the grades depend on the performance of the students. Second, the students were inducted at different levels (strongly inducted, lightly inducted, no induction at all) on how to solve the problem during the years, but always leaving open the option to do research and implement the algorithm they think is best in order to analyze the role of proactivity in this context. In this work we show the study and analysis of the last measure. We collected the number of similar approaches, grouped by five different categories (rule-based, traditional mapping, deep learning, deep learning extended and reinforcement learning) for the years 2016, 2017 and 2018, and we found out that the students aimed to and learned more sophisticated and complex solutions as the level of induction is reduced.

With the emergence of low cost 3D sensors, the focus is moving towards the recognition and scene understanding of tridimensional data. This kind of representation is really challenging in terms of computation, and it needs the development of new strategies and algorithms to be handled and interpreted.In this work, we propose NurbsNet, a novel approach for 3D object classification based on local similarities with free form surfaces modeled as Nurbs.The proposal has been tested in ModelNet10 and ModelNet40 with results that are promising with less training iterations than state-of-the-art methods and very low memory consumption.

Most state-of-the-art machine learning classifiers currently perform a direct classification of categories. Nonetheless, all the concepts are inherently connected by an inclusion dependency, creating a tree of relationships. For example, a car and a bus are elements included in the concept vehicle. Traditional approaches make use of finer grain concepts, thus discarding important data provided by these relational bounds. This paper describes different strategies to include this tree information in a deep learning pipeline. We compare our proposals, which produce multilevel classifications, with a direct classification baseline, outperforming it in each case. In addition, multilevel predictions could be further improved with different path correction strategies, which were also tested.

Un analisis en terminos educativos de la titulacion del grado en ingenieria en robotica pone de manifiesto la estrecha relacion entre las distintas areas de estudio abordadas durante su cursacion. Este hecho brinda la oportunidad de perfilar los conocimientos que se estudian en las distintas asignaturas (en particular: sistemas inteligentes, sistemas de percepcion y vision artificial) y aplicarlos mediante tecnicas de aprendizaje activo como el aprendizaje basado en proyectos. Asi, para este proyecto se ha propuesto el desarrollo de una aplicacion robotica asistencial para personas discapacitadas. Dada la complejidad de este problema, los estudiantes requerian de todos los conocimientos adquiridos para poder alcanzar una solucion viable. Ademas, la amplia variedad de posibles soluciones y problemas que surgen cuando se trabaja en escenarios reales ha resultado en una experiencia enriquecedora, involucrando a los estudiantes durante todo el desarrollo del proyecto. Esta participacion les ha proporcionado nuevas habilidades como autonomia, investigacion, planificacion, creatividad o analisis de datos, asi como los beneficios de trabajar en equipo, desarrollando habilidades comunicativas y tomando conciencia de las preocupaciones sociales, problemas clave en cualquier carrera de ingenieria. Asi pues, este estudio piloto ha evaluado el aprendizaje del alumnado cuando se enfrenta a un problema de la vida real, asi como la interrelacion entre asignaturas.

3D data sensors provide an enormous amount of information. It is necessary to develop efficient methods to manage this information under certain time, bandwidth or storage space requirements. In this work, we propose a 3D compression and decompression method. This method also allows the use of the compressed data for a registration process. First, points are selected and grouped, using a 3D-model based on planar surfaces. Next, we use a fast variant of Gaussian Mixture Models and an Expectation-Maximization algorithm to replace the points grouped in the previous step with a set of Gaussian distributions. These learned models can be used as features to find matches between two consecutive poses and apply 3D pose registration using RANSAC. Finally, the 3D map can be obtained by decompressing the models. (C) 2018 Elsevier B.V. All rights reserved.

In 2015, a new degree has been implemented at University of Alicante, degree of robotics. It is the first degree about robotics in Spain and one of the first in the world. Although there are several similar degrees related with robotics, this one consists in a mixture between industrial engineering and computer science. The implementation of this degree has several challenges to be addressed.

In several computer science degrees there are subjects related to computer vision or robotics. Although both computer vision and robotics can be taught using different approaches, we present here an approach more related with a computer science degree. In this paper, several examples of practical lesson are presented, which can serve as a basis for teaching those concepts in a computer science degree.

In 2015, a new degree has been implemented at University of Alicante, degree of robotics. It is the first degree about robotics in Spain and one of the first degrees in the world. Although there are several similar degrees related with robotics, this one consists in a mixture between industrial engineering and computer science. The implementation of this degree has several challenges to be addressed.In this paper, we present the methodology in the implementation and the contents of the programming subject in the degree. This implementation must be carried out keeping in mind the future subjects of the degree. Programming is a core skill for this kind of engineers but, by other hand, students have not got an acquired programming skill. So, we need to address the trade-off between implementing a traditional programming subject (limited by the limited number of related subjects) and obtaining the important programming skills. The paper shows the experience implementing the subject Programming 1 for the first year of this degree.For the practical sessions, two projects must be carried out. The first one is guided and introduces students to programming through the development of a simple navigation algorithm. It is guided and introduces students to programming. But in the second one we propose a computer vision project, introducing the students into the computer vision world. It is worth to say that there is a Computer Vision subject in the third course. We detailed the practical proposed project.This paper could serve as a guide for future robotics degrees.

RGB-D sensors are capable of providing 3D points (depth) together with color information associated with each point. These sensors suffer from different sources of noise. With some kinds of RGB-D sensors, it is possible to pre-process the color image before assigning the color information to the 3D data. However, with other kinds of sensors that is not possible: RGB-D data must be processed directly. In this paper, we compare different approaches for noise and artifacts reduction: Gaussian, mean and bilateral filter. These methods are time consuming when managing 3D data, which can be a problem with several real time applications. We propose new methods to accelerate the whole process and improve the quality of the color information using entropy information. Entropy provides a framework for speeding up the involved methods allowing certain data not to be processed if the entropy value of that data is over or under a given threshold. The experimental results provide a way to balance the quality and the acceleration of these methods. The current results show that our methods improve both the image quality and processing time, as compared to the original methods.

In several computer science degrees there are subjects related to computer vision or robotics. Although those subjects are usually related to industrial engineers, those areas are closely related to computer science, i.e., everyday more and more computer science students begin a master's degree or PhD in robotics. This paper could serve as a guided for future implementation of computer vision and robotics concepts for a computer science degree.

The use of 3D data in mobile robotics applications provides valuable information about the robot's environment. However usually the huge amount of 3D information is difficult to manage due to the fact that the robot storage system and computing capabilities are insufficient. Therefore, a data compression method is necessary to store and process this information while preserving as much information as possible.A few methods have been proposed to compress 3D information. Nevertheless, there does not exist a consistent public benchmark for comparing the results (compression level, distance reconstructed error, etc.) obtained with different methods. In this paper, we propose a dataset composed of a set of 3D point clouds with different structure and texture variability to evaluate the results obtained from 3D data compression methods. We also provide useful tools for comparing compression methods, using as a baseline the results obtained by existing relevant compression methods. (C) 2015 Elsevier B.V. All rights reserved.

In this article, we present a new framework oriented to teach Computer Vision related subjects called JavaVis. It is a computer vision library divided in three main areas: 2D package is featured for classical computer vision processing; 3D package, which includes a complete 3D geometric toolset, is used for 3D vision computing; Desktop package comprises a tool for graphic designing and testing of new algorithms. JavaVis is designed to be easy to use, both for launching and testing existing algorithms and for developing new ones. © 2013 Wiley Periodicals, Inc. Comput Appl Eng Educ 23:258–267, 2015; View this article online at wileyonlinelibrary.com/journal/cae; DOI 10.1002/cae.21594

The use of 3D data in mobile robotics provides valuable information about the robot's environment. Traditionally, stereo cameras have been used as a low-cost 3D sensor. However, the lack of precision and texture for some surfaces suggests that the use of other 3D sensors could be more suitable. In this work, we examine the use of two sensors: an infrared SR4000 and a Kinect camera. We use a combination of 3D data obtained by these cameras, along with features obtained from 2D images acquired from these cameras, using a Growing Neural Gas (GNG) network applied to the 3D data. The goal is to obtain a robust egomotion technique. The GNG network is used to reduce the camera error. To calculate the egomotion, we test two methods for 3D registration. One is based on an iterative closest points algorithm, and the other employs random sample consensus. Finally, a simultaneous localization and mapping method is applied to the complete sequence to reduce the global error. The error from each sensor and the mapping results from the proposed method are examined. (C) 2014 Elsevier Inc. All rights reserved.

Nowadays, there is an increasing number of robotic applications that need to act in real three-dimensional (3D) scenarios. In this paper we present a new mobile robotics orientated 3D registration method that improves previous Iterative Closest Points based solutions both in speed and accuracy. As an initial step, we perform a low cost computational method to obtain descriptions for 3D scenes planar surfaces. Then, from these descriptions we apply a force system in order to compute accurately and efficiently a six degrees of freedom egomotion. We describe the basis of our approach and demonstrate its validity with several experiments using different kinds of 3D sensors and different 3D real environments.