Abstract
This paper addresses the development of a novel zero-shot learning method for remote marine litter hyperspectral imaging data classification. The work consisted of using an airborne acquired marine litter hyperspectral imaging dataset that contains data about different plastic targets and other materials and assessing the viability of detecting and classifying plastic materials without knowing their exact spectral response in an unsupervised manner. The classification of the marine litter samples was divided into known and unknown classes, i.e., classes that were hidden from the dataset during the training phase. The obtained results show a marine litter automated detection for all the classes, including (in the worst case of an unknown class) a precision rate over 56% and an overall accuracy of 98.71%.

Abstract

Abstract
This paper addresses the development of a remote hyperspectral imaging system for detection and characterization of marine litter concentrations in an oceanic environment. The work performed in this paper is the following: (i) an in-situ characterization was conducted in an outdoor laboratory environment with the hyperspectral imaging system to obtain the spatial and spectral response of a batch of marine litter samples; (ii) a real dataset hyperspectral image acquisition was performed using manned and unmanned aerial platforms, of artificial targets composed of the material analyzed in the laboratory; (iii) comparison of the results (spatial and spectral response) obtained in laboratory conditions with the remote observation data acquired during the dataset flights; (iv) implementation of two different supervised machine learning methods, namely Random Forest (RF) and Support Vector Machines (SVM), for marine litter artificial target detection based on previous training. Obtained results show a marine litter automated detection capability with a 70–80% precision rate of detection in all three targets, compared to ground-truth pixels, as well as recall rates over 50%.

Abstract
Achieving persistent and reliable autonomy for mobile robots in challenging field mission scenarios is a long-time quest for the Robotics research community. Deep learning-based LIDAR odometry is attracting increasing research interest as a technological solution for the robot navigation problem and showing great potential for the task. In this work, an examination of the benefits of leveraging learning-based encoding representations of real-world data is provided. In addition, a broad perspective of emergent Deep Learning robust techniques to track motion and estimate scene structure for real-world applications is the focus of a deeper analysis and comprehensive comparison. Furthermore, existing Deep Learning approaches and techniques for point cloud odometry tasks are explored, and the main technological solutions are compared and discussed. Open challenges are also laid out for the reader, hopefully offering guidance to future researchers in their quest to apply deep learning to complex 3D non-matrix data to tackle localization and robot navigation problems.

Abstract