Robotics and autonomous systems have been key to space exploration and science and will continue to be instrumental.

In semi-supervised medical image segmentation, most previous works draw on the common assumption that higher entropy means higher uncertainty. In this paper, we investigate a novel method of estimating uncertainty. We observe that, when assigned different misclassification costs in a certain degree, if the segmentation result of a pixel becomes inconsistent, this pixel shows a relative uncertainty in its segmentation. Therefore, we present a new semi-supervised segmentation model, namely, conservative-radical network ( CoraNet in short) based on our uncertainty estimation and separate self-training strategy. In particular, our CoraNet model consists of three major components: a conservative-radical module (CRM), a certain region segmentation network (C-SN), and an uncertain region segmentation network (UC-SN) that could be alternatively trained in an end-to-end manner. We have extensively evaluated our method on various segmentation tasks with publicly available benchmark datasets, including CT pancreas, MR endocardium, and MR multi-structures segmentation on the ACDC dataset. Compared with the current state of the art, our CoraNet has demonstrated superior performance. In addition, we have also analyzed its connection with and difference from conventional methods of uncertainty estimation in semi-supervised medical image segmentation.

Reducing the risk of oil spill disasters is essential for protecting the environmentand reducing economic losses. Oil spill surveillance constitutes an important component ofoil spill disaster management. Advances in remote sensing technologies can help to identifyparties potentially responsible for pollution and to identify minor spills before they causewidespread damage. Due to the large number of sensors currently available for oil spillsurveillance, there is a need for a comprehensive overview and comparison of existingsensors. Specifically, this paper examines the characteristics and applications of differentsensors. A better understanding of the strengths and weaknesses of oil spill surveillancesensors will improve the operational use of these sensors for oil spill response andcontingency planning. Laser fluorosensors were found to be the best available sensor for oilspill detection since they not only detect and classify oil on all surfaces but also operate ineither the day or night. For example, the Scanning Laser Environmental AirborneFluorosensor (SLEAF) sensor was identified to be a valuable tool for oil spill surveillance.However, no single sensor was able to provide all information required for oil spillcontingency planning. Hence, combinations of sensors are currently used for oil spillsurveillance. Specifically, satellite sensors are used for preliminary oil spill assessmentwhile airborne sensors are used for detailed oil spill analysis. While satellite remote sensingis not suitable for tactical oil spill planning it can provide a synoptic coverage of theaffected area.

Starting from 2016, the raw Global Navigation Satellite System (GNSS) measurements can be extracted from the Android Nougat (or later) operating systems. Since then, GNSS smartphone positioning has been given much attention. A high number of related publications indicates the importance of the research in this field, as it has been doing in recent years. Due to the cost-effectiveness of the GNSS smartphones, they can be employed in a wide variety of applications such as cadastral surveys, mapping surveying applications, vehicle and pedestrian navigation and etc. However, there are still some challenges regarding the noisy smartphone GNSS observations, the environment effect and smartphone holding modes and the algorithm development part which restrict the users to achieve high-precision smartphone positioning. In this review paper, we overview the research works carried out in this field with a focus on the following aspects: first, to provide a review of fundamental work on raw smartphone observations and quality assessment of GNSS observations from major smart devices including Google Pixel 4, Google Pixel 5, Xiaomi Mi 8 and Samsung Ultra S20 in terms of their signal strengths and carrier-phase continuities, second, to describe the current state of smartphone positioning research field until most recently in 2021 and, last, to summarize major challenges and opportunities in this filed. Finally, the paper is concluded with some remarks as well as future research perspectives.