Abstract: Small drones are being utilized in monitoring, delivery of goods, public safety, and disaster management among other civil applications. Due to their sizes, capabilities, payload limitations, and limited flight time, it is not far-fetched to expect multiple networked and coordinated drones incorporated into the air traffic. In this paper, we describe a high-level architecture for the design of a collaborative aerial system that consists of drones with on-board sensors and embedded processing, sensing, coordination, and communication&networking capabilities. We present a multi-drone system consisting of quadrotors and demonstrate its potential in a disaster assistance scenario. Furthermore, we illustrate the challenges in the design of drone networks and present potential solutions based on the lessons we have learned so far.

Abstract: A convergence of trends is shifting the focus of quality assessment from compliance with system design goals to fulfillment of user needs or expectations in different contexts.

Abstract: This paper describes a design framework for TCPfriendly and media-friendly rate control algorithms for multimedia streaming applications. The idea of this framework is to start from TFRC’s (TCP-Friendly Rate Control) transmission rate and then alter this transmission rate so that it tracks the media characteristics of the stream (e.g., bitrate) or other application characteristics like the client buffer fill level. In this way, the media-friendly property of the algorithm is achieved. We give three rules that guide how the TFRC throughput should track the evolution of the stream’s media characteristics and remain TCPfriendly in the long term. We also present, as proof of concept, four simple media-friendly and TCP-friendly congestion control algorithms built using the aforementioned framework. These congestion control algorithms are better suited for multimedia streaming applications than traditional TCP congestion control or smooth congestion control algorithms like TFRC. We have performed evaluations of two of the four proposed media-friendly and TCP-friendly congestion control algorithms under various network conditions and validated that they represent viable transport solutions, better than TFRC, for variable bitrate video streams. More specifically, our two media-friendly and TCPfriendly congestion control algorithms maintained a TCP-friendly throughput in the long term in all experiments and avoided an empty buffer at the client side in situations when TFRC could not achieve this.

Abstract: This paper proposes and evaluates a modular architecture of an autonomous unmanned aerial vehicle (UAV) system for search and rescue missions. Multiple multicopters are coordinated using a distributed control system. The system is implemented in the Robot Operating System (ROS) and is capable of providing a real-time video stream from a UAV to one or more base stations using a wireless communications infrastructure. The system supports a heterogeneous set of UAVs and camera sensors. If necessary, an operator can interfere and reduce the autonomy. The system has been tested in an outdoor mission serving as a proof of concept. Some insights from these tests are described in the paper.

Abstract: Multimedia content delivery and real-time streaming over the top of the existing infrastructure is nowadays part and parcel of every media ecosystem thanks to open standards and the adoption of the Hypertext Transfer Protocol (HTTP) as its primary mean for transportation. Hardware encoder manufacturers have adopted their product lines to support the dynamic adaptive streaming over HTTP but suffer from the inflexibility to provide scalability on demand, specifically for event-based live services that are only offered for a limited period of time. The cloud computing paradigm allows for this kind of flexibility and provide the necessary elasticity in order to easily scale with the demand required for such use case scenarios. In this paper we describe bitcodin, our transcoding and streaming-as-as-ervice platform based on open standards (i.e., MPEG-DASH) which is deployed on standard cloud and content delivery infrastructures to enable high-quality streaming to heterogeneous clients. It is currently deployed for video on demand, 24/7 live, and event-based live services using bitdash, our adaptive client framework.

Abstract: Our approach to automatically separating compound figures appearing in biomedical articles is split into two image processing algorithms: one is based on detecting separator edges, and the other tries to identify background bands separating subgures. Only one algorithm is applied to a given image, according to the prediction of a binary classifier trained to distinguish graphical illustrations from other images in biomedical articles. Our submission to the ImageCLEF 2015 compound figure separation task achieved an accuracy of 49% on the provided test set of about 3400 compound images. This stays clearly behind the best submission of other participants (85% accuracy), but is by an order of magnitude faster than other approaches reported in the literature.

Abstract: We propose a new interface that facilitates content navigation in videos on devices with touchscreen interaction. This interface allows both coarse-grained and fine-grained navigation in an intuitive way and enables better performance when used to locate specific scenes in videos. We implemented this interface on a 5.5-inch smartphone and tested it with 24 users.Our results show that for video navigation tasks the proposed interface significantly outperforms the seeker-bar interface, commonly used with video players on mobile devices. Moreover, we found that the interaction concept of the Scrubbing Wheel has a much lower perceived workload than the widely used seeker-bar, and is the preferred tool to locate scenes in videos for all tested users in our study.

Abstract: Social networks have become ubiquitous and with these new possible ways for social communication and experiencing multimedia together the traditional TV scenario drifts more and more towards a distributed social experience. Asynchronism in the multimedia playback of the users may have a significant impact on the acceptability of systems providing the distributed multimedia experience. The synchronization needed in such systems is called Inter-Destination Multimedia Synchronization. In this paper we propose a demo that implements IDMS by the means of our self-organized and distributed approach assisted by pull-based streaming. We also provide a video of the planned demonstration and provide the mobile application as open source licensed under the GNU LGPL.

Abstract: Modern-age technology enables us to consume multimedia for enjoyment and as a social experience. The traditional way to consume multimedia together (e.g., with family or friends in the living room) is being superseded by a location-independent scenario where geographically distributed users consume the same content while having a real-time communication channel among each other. Inter-Destination Multimedia Synchronization (IDMS) is the tool of choice in order to enable users a high-quality multimedia experience. In this paper, we investigate the influence of asynchronism when consuming multimedia content together while being geographically distributed. In particular, we adopt the concept of human computation and developed a reaction game which we used to conduct a crowdsourced subjective quality assessment in order to evaluate a threshold for multimedia synchronization within an IDMS scenario. Our results show a significant decrease in overall Quality of Experience (QOE) at an asynchronism level of 750ms. At the same time, we were able to show that asynchronism at a level of 400ms does not have significant differences regarding the QoE when compared to the synchronous reference case.

Abstract: Multimedia delivery systems and protocols usually assume end-to-end connections and low delivery delays between multimedia sources and consumers. However, neither of these two properties can always be achieved in hastily formed networks for emergency response operations. In particular, disruptions may break end-to-end connections, which makes it impossible to deliver multimedia content instantly. This work presents a multimedia delivery system that can operate in disrupted networks and hence may help improve the situational awareness in emergency response operations. The multimedia delivery system is based on HTTP adaptive streaming (HAS) and uses a modified version of HTTP which is able to deliver data in partitioned networks. The multimedia delivery system is evaluated in a realistic emergency response scenario.

Abstract: In medical endoscopy more and more surgeons record videos of their interventions in a long-term storage archive for later retrieval. In order to allow content-based search in such endoscopic video archives, the video data needs to be indexed first. However, even the very basic step of content-based indexing, namely content segmentation, is already very challenging due to the special characteristics of such video data. Therefore, we propose to use instrument classification to enable semantic segmentation of laparoscopic videos. In this paper, we evaluate the performance of such an instrument classification approach. Our results show satisfying performance for all instruments used in our evaluation.

Abstract: Social media becomes a vital part in our daily communication practice, creating a huge amount of data and covering different real-world situations. Currently, there is a tendency in making use of social media during emergency management and response. Most of this effort is performed by a huge number of volunteers browsing through social media data and preparing maps that can be used by professional first responders. Automatic analysis approaches are needed to directly support the response teams in monitoring and also understanding the evolution of facts in social media during an emergency situation. In this paper, we investigate the problem of real-time sub-events identification in social media data (i.e., Twitter, Flickr and YouTube) during emergencies. A processing framework is presented serving to generate situational reports/summaries from social media data. This framework relies in particular on online indexing and online clustering of media data streams. Online indexing aims at tracking the relevant vocabulary to capture the evolution of sub-events over time. Online clustering, on the other hand, is used to detect and update the set of sub-events using the indices built during online indexing. To evaluate the framework, social media data related to Hurricane Sandy 2012 was collected and used in a series of experiments. In particular some online indexing methods have been tested against a proposed method to show their suitability. Moreover, the quality of online clustering has been studied using standard clustering indices. Overall the framework provides a great opportunity for supporting emergency responders as demonstrated in real-world emergency exercises.

Abstract: In crisis situations like riots, earthquakes, storms, etc. information plays a central role in the process of organizing interventions and decision making. Due to their increasing use during crises, social media (SM) represents a valuable source of information that could help obtain a full picture of people needs and concerns. In this chapter, we highlight the importance of SM networks in crisis management (CM) to show how information is propagated through. The chapter also summarizes the current state of research related to information propagation in SM networks during crises. In particular three classes of information propagation research categories are identified: network analysis and community detection, role and topic-oriented information propagation, and infrastructure-oriented information propagation. The chapter describes an analysis framework that deals with structural information propagation for crisis management purposes. Structural propagation is about broadcasting specific information obtained from social media networks to targeted sinks/receivers/hubs like emergency agencies, police department, fire department, etc. Specifically, the framework aims to identify the discussion topics, known as sub-events , related to a crisis (event) from SM contents. A brief description of techniques used to detect topics and the way those topics can be used in structural information propagation are presented.

Abstract: Streaming multimedia over the Internet is omnipresent but still in its infancy, specifically when it comes to the adaptation based on bandwidth/throughput measurements, clients competing for limited/shared bandwidth, and the presence of a caching infrastructure. In this paper we present a buffer-based adaptation logic in combination with a toolset of client metrics to compensate for erroneous adaptation decisions. These erroneous adaptation decisions are due to insufficient network information available at the client and issues introduced when multiple clients compete for limited/shared bandwidth and/or when caches are deployed. Our metrics enable the detection of oscillations on the client - in contrast to server-based approaches - and provide an effective compensation mechanism. We evaluate the proposed adaptation logic, which incorporates the oscillation detection and compensation method, and compare it against a throughput-based adaptation logic for scenarios comprising competing clients with and without caching enabled. In anticipation of the results, we show how the presented metrics detect oscillation periods and how such undesirable situations can be compensated while increasing the effective media throughput of the clients.

Abstract: The focus of our research is on improving mobile image and video browsing interfaces. To get a better idea about real world mobile photo and video scenarios and to base our research on real world numbers we performed a survey of photo and video usage on smartphones and tablets. In an online survey we asked 215 participants of the German speaking region about their mobile image collections, their usage patterns, and their motives and intentions when capturing photos. Our results show, among other things, that users store considerable more photos on smartphones than on tablets, that the majority of our participants use their smartphone as primary camera and that users are unlikely to organize their photos on their mobile devices in any way. Moreover, the most popular motives are people, holiday photos, events, and landscapes. Furthermore, it is more popular to capture photos for private than for sharing purposes. We also report about various correlation hypothesis that we tested in the gathered data.