Computer Science

In recent years, several studies have shown that the canine retina model offers important insight for our understanding of human retinal diseases. Several therapies developed to treat blindness in such models have already moved onto human clinical trials, with more currently under development [1]. Optical coherence tomography (OCT) offers a high resolution imaging modality for performing in-vivo analysis of the retinal layers. However, existing algorithms for automatically segmenting and analyzing such data have been mostly focused on the human retina. As a result, canine retinal images are often still being analyzed using manual segmentations, which is a slow and laborious task. In this work, we propose a method for automatically segmenting 5 boundaries in canine retinal OCT. The algorithm employs the position relationships between different boundaries to adaptively enhance the gradient map. A region growing algorithm is then used on the enhanced gradient maps to find the five boundaries separately. The automatic segmentation was compared against manual segmentations showing an average absolute error of 5.82 ± 4.02 microns.

This paper describes a smart parking sensing and information system that disseminates parking availability information to public users in a cost-effective and efficient manner. The hardware framework of the system is built on advanced wireless sensor networks and cloud service over the Internet, and the system is highly scalable. The parking information provided to the users is set in the form of occupancy rates and expected cruising time. Both are obtained from an analytical algorithm that processes historical and real-time data and are then visualized in a color theme. The entire parking system is deployed and extensively evaluated at Stanford University, California, Parking Structure 1.

In this study, we leverage the fusion of edge computing, artificial intelligence (AI) methods, and facilities provided by B5G to build a heterogeneous set of AI techniques for COVID-19 outbreak prediction. Advancement in the areas of AI, edge computing, the Internet of Things (IoT), and fast communication networks provided by beyond 5G (B5G) networks has opened doors for new possibilities by fusing these technologies and techniques. In a pandemic outbreak, such as COVID-19, the need for rapid analysis, decision making, and prediction of future trends becomes paramount. On a global map, the distributed processing and analysis of data at the source is now possible and much more efficient. With the features provided by B5G, such as low latency, larger area coverage, higher data rate, and realtime communication, building new intelligent and efficient frameworks is becoming easier. In this study, our aim is to achieve higher accuracy in prediction by fusing multiple AI methods and leveraging the B5G communication architecture. We propose a distributed architecture for training AI methods on edge devices, with the results of edge-trained models then propagated to a central cloud AI method, which then combines all the received edge-trained models into a global and final prediction model. The experimental results of five countries (United States, India, Italy, Bangladesh, and Saudi Arabia) show that the proposed distributed AI on edges can predict COVID-19 outbreak better than that of each individual AI method in terms of correlation coefficient scores.

In recent years, the advent of car navigation systems has laid the ground for an entirely new industry sector, consisting of map producers, car/ personal/ smart phone navigation manufacturers, and service providers. It has probably gone unnoticed that navigation systems mark a major change in the way we use maps. Partially, they are still just a replacement for traditional maps, providing a means to store and visualize a representation of the environment. In contrast to the traditional use of maps, however, navigation systems perform computations using the map's data structures, such as shortest route, map matching, and route guidance. That is, from an abstract point of view, part of the map is made for machine use only – the user has no direct access to it but rather is only presented the outcome of the computations.

This paper introduces a multi-view stereo matcher that generates depth in real-time from a monocular video stream of a static scene. A key feature of our processing pipeline is that it estimates global camera gain changes in the feature tracking stage and efficiently compensates for these in the stereo stage without impacting the real-time performance. This is very important for outdoor applications where the brightness range often far exceeds the dynamic range of the camera. Real-time performance is achieved by leveraging the processing power of the graphics processing unit (GPU) in addition to the CPU. We demonstrate the effectiveness of our approach on videos of urban scenes recorded by a vehicle-mounted camera with auto-gain enabled.

From September 12th to 17th, 2010, the Dagstuhl Seminar 10371 ``Dynamic Maps '' was held in Schloss Dagstuhl~--~Leibniz Center for Informatics. During the seminar, several participants presented their current research, and ongoing work and open problems were discussed. Abstracts of the presentations given during the seminar as well as abstracts of seminar results and ideas are put together in this paper. The first section describes the seminar topics and goals in general. Links to extended abstracts or full papers are provided, if available.