Abstract: The Scalable Video Coding (SVC) extension of the H.264/AVC (AVC) video coding standard features spatial, quality and temporal scalability. Backwards compatibility with legacy decoding devices is maintained through an H.264/AVC compliant base layer, which represents the lowest quality of an SVC bit-stream. However, it is often desirable to also provide the higher quality layers to legacy H.264/AVC devices. This is achieved by a process commonly known as “bit-stream rewriting”, which allows for an efficient SVC to AVC conversion by exploiting the similarities of the two codecs. This paper introduces an improved version of the existing JSVM reference software rewriter (JSVM-rewriter). The improvements include a better run-time performance through parallel processing, as well as applicability in streaming scenarios. A detailed evaluation provides performance measurements for the improved rewriter and compares it to the existing JSVM-rewriter. The evaluation shows that notable performance improvements can be achieved using the presented approach. The paper concludes on how the rewriter could be further improved.

Abstract: This paper describes the CELTIC project on "Scalable Video Coding Impact on Networks" with the focus of designing a streaming system based on the Scalable Video Coding extension of the H.264/AVC standard. The system is designed to cope with streaming scenarios that can be classified in four use cases: session handover, network congestion, receiver heterogeneity and user driven adaptation. A complete overview of the architecture of the system is given. Two demonstration scenarios are described in detail, which point out the advantages of scalable video coding compared to single layer approaches in multimedia transmission and adaptation scenarios. A concluding section summarizes the work and provides an outlook to future work items.

Abstract: This paper compares handhelds based on the iPhone and Android operating systems in multimedia streaming scenarios. We simulate typical Internet network impairments, i.e. packet delay and packet loss, and evaluate their effects on the end devices. Additional evaluations include bandwidth overhead in icted by the different streaming approaches and traffic shape and fairness when both handhelds consume media simultaneously. Based on the quantitative evaluation, both approaches show weaknesses and strengths. A final qualitative discussion points out additional advantages for the streaming approach implemented in the iPhone operating system.

Abstract: H.264/SVC enables runtime-efficient scalability in the spatial, temporal and fidelity dimension. Existing adaptation mechanisms facilitate this to automatically adapt the H.264/ SVC stream to the current usage environment without any user interaction. This paper argues that the Quality of Experience (QoE) of the end user can be enhanced by enabling him to manually adjust the adaptation if he wishes to do so. An approach which enables this is presented and evaluated. It is shown that by facilitating this approach an increased QoE is provided compared to automatic adaptation approaches. Finally, future work indicates the next steps in order to implement this approach.

Abstract: In this paper, a novel low level feature for content based image retrieval is presented. The proposed feature structure combines color and spatial color distribution information. The combination of these two features in one vector classifies the proposed descriptor to the family of Composite Descriptors. In order to extract the color information, a fuzzy system is being used, which is mapping the number of colors that are included in the image into a custom palette of 8 colors. The way by which the vector of the proposed descriptor is being formed, describes the color spatial information contained in images. To be applicable in the design of large image databases, the proposed descriptor is compact, requiring only 48 bytes per image. Experiments presented in this paper demonstrate the effectiveness of the proposed technique especially for Hand-Drawn Sketches.

Abstract: Recent advances in the fields of digital photography, networking and computing, have made it easier than ever for users to store and share photographs. However without sufficient metadata, e.g., in the form of tags, photos are difficult to find and organize. In this paper, we describe a system that recommends tags for image annotation. We postulate that the use of low-level global visual features can improve the quality of the tag recommendation process when compared to a baseline statistical method based on tag co-occurrence. We present results from experiments conducted using photos and metadata sourced from the Flickr photo website that suggest that the use of visual features improves the mean average precision (MAP) of the system and increases the system's ability to suggest different tags, therefore justifying the associated increase in complexity.

Abstract: Why do people take pictures? While substantial work has been done on the intentions of people producing photos in specific domains (e.g. Flickr users or mobile phone cameras) the general case - arbitrary users taking photos with arbitrary devices - has not yet been investigated in detail. In this short paper, we present an exploratory study on the intentions and goals of people taking digital photos. 40 different specific photo taking situations of 10 people where investigated. We classified the situations and reflected them into context of existing work to investigate the applicability and usefulness of classifications existing for specific domains. Our findings show that applicability of existing models for small domains is limited and they indicate the need of a generalized taxonomy. We further identify hypotheses and research questions for future work in this area.

Abstract: Resolution of digital images is on the rise, but screens of mobile devices are still small. Therefore, image adaptation and especially image re-targeting for browsing images is still a challenging research topic. In this short paper we report work in progress on a social interactive game that can be used to identify meaningful portions of images. Based on preliminary evaluation we propose that these areas, found by our game, should be retained in an image re-targeting process.

Abstract: Searching for images on the web is still an open problem. While multiple approaches have been presented, there has been surprisingly little work on the actual goals and intentions of users. In this poster we present our classification scheme for user goals in image search and describe our ongoing work focusing on identification and classification of user intentions during image search tasks.

Abstract: This paper proposes a novel combination of arti- ficial intelligence planning and other techniques for improv- ing decision-making in the context of multi-step multime- dia content adaptation. In particular, it describes a method that allows decision-making (selecting the adaptation to perform) in situations where third-party pluggable multi- media conversion modules are involved and the multime- dia adaptation planner does not know their exact adapta- tion capabilities. In this approach, the multimedia adapta- tion planner module is only responsible for a part of the required decisions; the pluggable modules make additional decisions based on different criteria. We demonstrate that partial decision-making is not only attainable, but also in- troduces advantages with respect to a system in which these conversion modules are not capable of providing additional decisions. This means that transferring decisions from the multi-step multimedia adaptation planner to the pluggable conversion modules increases the flexibility of the adapta- tion. Moreover, by allowing conversion modules to be only partially described, the range of problems that these modules can address increases, while significantly decreasing both the description length of the adaptation capabilities and the planning decision time. Finally, we specify the conditions under which knowing the partial adaptation capabilities of a set of conversion modules will be enough to compute a proper adaptation plan.

Abstract: Recent work in TCP video streaming indicates that multimedia streaming via TCP provides satisfactory performance when the achievable TCP throughput is approximately twice the media bit rate. However, these conditions may not be achievable on the Internet, e.g., when the delivery path offers insufficient bandwidth or becomes congested due to competing traffic. Therefore, adaptive streaming for videos over TCP is required and a number of rate-control algorithms for video streaming have been proposed and evaluated in the literature. In this paper, we evaluate and compare three existing rate-control algorithms for TCP streaming in terms of the (PSNR) quality of the delivered video and in terms of the timeliness of delivery. The contribution of the paper is that, to the best of our knowledge, this is the first evaluation of TCP-based streaming in an Internet-like setting making use of the scalability features of the H.264/SVC video codec. Two simple bandwidth estimation algorithms and a priority-/deadline-driven approach are described to adapt the bit rates of, and transmit, the H.264/SVC video in a rate-distortion optimal manner. The results indicate that the three algorithms perform robustly in terms of video quality and timely delivery, both on under-provisioned links and in case of competing TCP flows. The priority-/deadline-driven technique is even more stable in terms of packet delays and jitter; thus, client buffers can be dimensioned more easily.

Abstract: TCP-based video streaming encounters difficulties in unreliable networks with unanticipated packet loss. In combination with high round trip times, the effective throughput deteriorates rapidly and TCP connection resets or stalls may occur. In this paper, we propose a client-driven video transmission scheme which utilizes multiple HTTP/TCP streams. The scheme is largely insensitive to unanticipated packet loss and thereby reduces throughput fluctuations. Since it is based on HTTP, the scheme can easily be deployed in existing network infrastructures. It fosters scalability on the server side by shifting complexity from the server to the clients. Certain features of request-response schemes allow maintaining fairness, despite of using multiple HTTP streams. Making use of TCP, the scheme inherently adapts to congested network links.

Abstract: This thesis investigates mechanisms and applications for in-network adaptation of scalable video bit streams based on the recent H.264/Scalable Video Coding (SVC) standard. In-network adaptation refers to the adaptation of a video stream by a network element during the stream's transport through the network. The advantages of performing adaptation directly in the network are the availability of local monitoring data and a higher responsiveness according to the current networking conditions. In contrast to previous work in this field, this thesis focuses on the feasibility and realization of in-network adaptation on existing home router platforms. In this context this thesis addresses the following six research objectives. Initially, the relevant transport mechanisms for H.264/SVC and their implications on in-network adaptation (1) were analysed. In the context of this work three different Linux-based router platforms which cover a representative range of residential router devices were used as a basis for further studies and evaluations. In general these platforms can be characterized by rather modest processing capabilities and networking performance. The hardware limitations were identified and quantified in evaluations (2) using both different benchmarks and real network traffic. The offered processing power and memory throughput are roughly 10 to 100 times lower than those of a modern desktop PC. Although their application-layer networking performance is not that low, all platforms fail in fully utilizing their nominal link capacities of 100 and 1000 Mbps, respectively. Based on the known limitations the thesis proposes a stateful, packet-based adaptation mechanism for adapting scalable video bit streams (3). The approach utilizes the RTP payload format for H.264/SVC and represents a light-weight approach for in-network adaptation on the application layer. It further meets the important requirements towards a media-aware network element (MANE) to be signaling aware and to operate statefully. The mechanism was integrated in a proxy service which was deployed on all of the three platforms to prove its feasibility. Experimental evaluations with different video bit streams in standard-definition quality demonstrate the scalability of the approach (4). The results indicate that the proxy service is able to adapt up to 16 concurrent video streams depending on the platform and video bit stream. On two of the three evaluated platforms the proposed approach even allows to handle and to adapt video streams in high-definition quality at bit rates around 15 Mbps. In addition to the proposed H.264/SVC-specific adaptation mechanism, also the applicability of generic metadata-driven adaptation on home router platforms was investigated. In particular, a proof-of-concept study of an XML-metadata-driven approach based on the MPEG-21 generic Bitstream Syntax Description (gBSD) was conducted on the platforms (5). In contrast to former evaluations that have been done on PC-based platforms, the obtained results indicate that the use of this generic adaptation cannot be recommended on such resource limited network devices. The benefits of using in-network adaptation on home router platforms are finally demonstrated in the context of high-definition streaming over IEEE 802.11 wireless networks (6). Monitoring information regarding the queueing delay, which is obviously available exclusively on the router, is used to control the adaptation of the video according to the varying throughput of the wireless link. This allows to react timely to changing conditions particularly in the case of mobile clients.

Abstract: The distribution of layered content over peer-to-peer networks becomes more important today as the users are consuming the content on terminals with various display capabilities and different network connections. For single-layer content distribution, the piece-picking algorithm only needs to ensure that content pieces are downloaded in time for display. When layered content is distributed over a peer-to-peer network, the piece-picking algorithm needs to be modified to ensure that the best possible quality is displayed while all desired pieces still have to be received before their deadline expires. In this paper, the piece-picking problem for layered content is analyzed and a number of piece-picking algorithms for layered content based on the solutions for the knapsack problem are presented. Furthermore, an evaluation of these algorithms is performed and possible applications are discussed.

Abstract: We introduce an easy-to-use video browsing tool which assists users in getting a quick overview of videos as well as in finding segments of interest. It provides a parallel and a tree-based view for browsing the content of videos -- or even video collections -- in a hierarchical, non-sequential manner. The tool has a plug-in architecture and can be extended both by further presentation methods and by video analysis algorithms.

Abstract: We present an algorithm for video scene detection based on the identification of recurring motion sequences within a video stream. The motion information is extracted in the compressed domain of H.264/AVC videos, no full decoding of the video stream is needed. Based on the motion information our algorithm identifies sequences of adjacent frames with similar motion. Throughout all identified motion sequences we are searching for recurring patterns of similar ones. The most recurring pattern is used for the segmentation of the video stream into scenes. The evaluation shows promising results.

Abstract: Peer-to-Peer systems are nowadays a very popular solution for multimedia distribution, as they provide significant cost benefits compared with traditional server-client distribution. Additionally, the distribution of scalable content enables the consumption of the content in a quality suited for the available bandwidth and the capabilities of the end-user devices. Thus, the distribution of scalable content over Peer-to-Peer networks is a very actual research topic. This paper presents an architecture for the distribution of scalable content in a fully distributed Peer-to-Peer network. The architectural description includes how the scalable layers of the content are mapped to the pieces distributed in the Peer-to-Peer system and detailed descriptions of the producer- and consumer-site architecture of the system. The presented system is to our knowledge the first open-source Peer-to-Peer network with full Scalable Video Coding support.

Abstract: This paper proposes a novel virtual Content-Aware Network (CAN) layer as a part of a full layered architecture, focused, but not limited to, on multimedia distribution with Quality of Services (QoS) assurance. The overall system is based on a flexible cooperation between providers, operators and end-users, enabling users to access the offered multimedia services in various contexts and also to become private content providers. The paper introduces the main concepts and architecture for the main virtual network layer (i.e., CAN), exposing its role and interfaces among overall system layers. This work is a part of the starting effort inside of a new European FP7 ICT research project, ALICANTE.