Search

Face recognition is a technology with great potential in the field of robotics, due to its prominent role in human-robot interaction (HRI). This interaction is a keystone for the successful deployment of robots in areas requiring a customized assistance like education and healthcare, or assisting humans in everyday tasks. These unconstrained environments present additional difficulties for face recognition, extreme head pose variability being one of the most challenging. In this paper, we address this issue and make a fourfold contribution. First, it has been designed a tool for gathering an uniform distribution of head pose images from a person, which has been used to collect a new dataset of faces, both presented in this work. Then, the dataset has served as a testbed for analyzing the detrimental effects this problem has on a number of state-of-the-art methods, showing their decreased effectiveness outside a limited range of poses. Finally, we propose an optimization method to mitigate said negative effects by considering key pose samples in the recognition system&rsquo, s set of known faces. The conducted experiments demonstrate that this optimized set of poses significantly improves the performance of a state-of-the-art, cutting-edge system based on Multitask Cascaded Convolutional Neural Networks (MTCNNs) and ArcFace.

Jupyter notebooks are recently emerging as a valuable pedagogical resource in academy, being adopted in educational institutions worldwide. This is mainly due to their ability to combine the expressiveness of traditional explanations from textbooks, with the interaction capabilities of software applications, which provides numerous benefits for both students and lecturers of different fields. One of the areas that could benefit from their adoption is such of mobile robotics, whose recent popularity has resulted in an increasing demand of trained practitioners with a solid theoretical and practical background. Therefore, there is a need of high quality learning materials adapted to modern tools and methodologies. With that in mind, this work explores how the introduction of Jupyter notebooks in undergraduate mobile robotic courses contributes to improve both teaching and learning experiences. For that, we first present a series of (publicly available) educational notebooks encompassing a variety of topics relevant for robotics, with a particular emphasis in the study of mobile robots and commonly used sensors. Those documents have been built from the ground up to take advantage of the Jupyter Notebook framework, bridging the typical gap between theoretical frame and interactive code. We also present a case study describing the notebooks usage in undergraduate courses at University of M&aacute, laga, including a discussion on the promising results and findings obtained.