Abstract: In the past decade we observed the transition from push-based, fully managed media streaming to pull-based, unmanaged adaptive HTTP streaming thanks to enhancements in media compression, network capacity, and client capabilities. Adaptive media players, specifically their algorithms, have been subject to research for a long time and lead to various approaches documented in the literature. In the past years we witnessed more and more commercial deployments taking into account findings presented in scientific papers but a quantitative evaluation and assessments of its performance is missing. In this paper, we propose means for the automated performance evaluation of commercially deployed adaptive media players with respect to i) objective, well-known metrics, such as bitrate, stalls, startup delay and ii) derived/calculated metrics (instability, inefficiency, average bitrate) previously proposed in the literature. Additionally, we propose a new metric (Bandwidth index) to measure the effectiveness of bandwidth utilization and together with existing QoE models for adaptive HTTP streaming (focusing on stalls, startup delay) we demonstrate its usefulness in this domain.

Abstract: In hospitals all around the world, medical multimedia information systems have gained high importance over the last few years. One of the reasons is that an increasing number of interventions are performed in a minimally invasive way. These endoscopic inspections and surgeries are performed with a tiny camera -- the endoscope -- which produces a video signal that is used to control the intervention. Apart from the viewing purpose, the video signal is also used for automatic content analysis during the intervention as well as for post-surgical usage, such as communicating operation techniques, planning future interventions, and medical forensics. Another reason is video documentation, which is even enforced by law in some countries. The problem, however, is the sheer amount of unstructured medical videos that are added to the multimedia archive on a daily basis. Without proper management and a multimedia information system, the medical videos cannot be used efficiently for post-surgical scenarios. It is therefore already foreseeable that medical multimedia information systems will gain even more attraction in the next few years. In this tutorial we will introduce the audience to this challenging new field, describe the domain-specific characteristics and challenges of medical multimedia data, introduce related use cases, and talk about existing works -- contributed by the medical imaging and robotics community, but also already partly from the multimedia community -- as well as the many open issues and challenges that bear high research potential.

Abstract: This extended demo paper summarizes our interface used for the Video Browser Showdown (VBS) 2017 competition, where visual and textual known-item search (KIS) tasks, as well as ad-hoc video search (AVS) tasks in a 600-h video archive need to be solved interactively. To this end, we propose a very flexible distributed video search system that combines many ideas of related work in a novel and collaborative way, such that several users can work together and explore the video archive in a complementary manner. The main interface is a perspective Feature Map, which shows keyframes of shots arranged according to a selected content similarity feature (e.g., color, motion, semantic concepts, etc.). This Feature Map is accompanied by additional views, which allow users to search and filter according to a particular content feature. For collaboration of several users we provide a cooperative heatmap that shows a synchronized view of inspection actions of all users. Moreover, we use collaborative re-ranking of shots (in specific views) based on retrieved results of other users.

Abstract: Due to increasing possibilities to create digital video, we are facing the emergence of large video archives that are made accessible either online or offline. Though a lot of research has been spent on video retrieval tools and methods, which allow for automatic search in videos, still the performance of automatic video retrieval is far from optimal. At the same time, the organization of personal data is receiving increasing research attention due to the challenges that are faced in gathering, enriching, searching and visualizing this data. Given the increasing quantities of personal data being gathered by individuals, the concept of a heterogeneous personal digital libraries of rich multimedia and sensory content for every individual is becoming a reality. Despite the differences between video archives and personal lifelogging libraries, we are facing very similar challenges when accessing these multimedia repositories. For example, users will struggle to find the information they are looking for in either collection if they are not able to formulate their search needs through a query. In this tutorial we discussed (i) proposed solutions for improved video & lifelog content navigation, (ii) typical interaction of content-based querying features, and (iii) advanced content visualization methods. Moreover, we discussed and demonstrate interactive video & lifelog search systems and ways to evaluate their performance.

Abstract: Currently, we witness dramatically increasing interest in immersive media technologies like Virtual Reality (VR), particularly in omnidirectional video (OV) streaming. Omnidirectional (also called 360-degree) videos are panoramic spherical videos in which the user can look around during playback and which therefore can be understood as hybrids between traditional movie streaming and interactive VR worlds. Unfortunately, streaming this kind of content is extremely bandwidth intensive (compared to traditional 2D video) and therefore, Quality of Experience (QoE) tends to deteriorate significantly in absence of continuous optimal bandwidth conditions. In this paper, we present a first approach towards subjective QoE assessment for omnidirectional video (OV) streaming. We present the results of a lab study on the QoE impact of stalling in the context of OV streaming using head-mounted displays (HMDs). Our findings show that subjective testing for immersive media like OV is not trivial, with even simple cases like stalling leading to unexpected results. After a discussion of characteristic pitfalls and lessons learned, we provide a a set of recommendations for upcoming OV assessment studies.

Abstract: This paper describes our approach used for the fully automatic and manually assisted Ad-hoc Video Search (AVS) task for TRECVID 2017. We focus on the combination of different convolutional neural network models and query optimization. Each of this model focus on a specific query part, which could be, e.g., location, objects, or the wide-ranging ImageNet classes. All classification results are collected in different combinations in Lucene indixes. For the manually assisted run we use a junk filter and different query optimization methods.

Abstract: Over the last decades, progress in medical technology and imaging technology enabled the technique of minimally invasive surgery. In addition, multimedia technologies allow for retrospective analyses of surgeries. The accumulated videos and images allow for a speed-up in documentation, easier medical case assessment across surgeons, training young surgeons, as well as they find the usage in medical research. Considering a surgery lasting for hours of routine work, surgeons only need to see short video segments of interest to assess a case. Surgeons do not have the time to manually extract video sequences of their surgeries from their big multimedia databases as they do not have the resources for this time-consuming task. The thesis deals with the questions of how to semantically classify video frames using Convolutional Neural Networks into different semantic concepts of surgical actions and anatomical structures. In order to achieve this goal, the capabilities of predefined CNN architectures and transfer learning in the laparoscopic video domain are investigated. The results are expected to improve by domain-specific adaptation of the CNN input layers, i.e. by fusion of the image with motion and relevance information. Finally, the thesis investigates to what extent surgeons' needs are covered with the proposed extraction of relevant scenes.

Abstract: The field of similarity based image retrieval has experienced a game changer lately. Hand crafted image features have been vastly outperformed by machine learning based approaches. Deep learning methods are very good at finding optimal features for a domain, given enough data is available to learn from. However, hand crafted features are still means to an end in domains, where the data either is not freely available, i.e. because it violates privacy, where there are commercial concerns, or where it cannot be transmitted, i.e. due to bandwidth limitations. Moreover, we have to rely on hand crafted methods whenever neural networks cannot be trained effectively, e.g. if there is not enough training data. In this paper, we investigate a particular approach to combine hand crafted features and deep learning to (i) achieve early fusion of off the shelf handcrafted global image features and (ii) reduce the overall number of dimensions to combine both worlds. This method allows for fast image retrieval in domains, where training data is sparse.

Abstract: In the last decade, advances in endoscopic surgery resulted in vast amounts of video data which is used for documentation, analysis, and education purposes. In order to find video scenes relevant for aforementioned purposes, physicians manually search and annotate hours of endoscopic surgery videos. This process is tedious and time-consuming, thus motivating the (semi-)automatic annotation of such surgery videos. In this work, we want to investigate whether the single-frame model for semantic surgery shot classification is feasible and useful in practice. We approach this problem by further training of AlexNet, an already pre-trained CNN architecture. Thus, we are able to transfer knowledge gathered from the Imagenet database to the medical use case of shot classification in endoscopic surgery videos. We annotate hours of endoscopic surgery videos for training and testing data. Our results imply that the CNN-based single-frame classification approach is able to provide useful suggestions to medical experts while annotating video scenes. Hence, the annotation process is consequently improved. Future work shall consider the evaluation of more sophisticated classification methods incorporating the temporal video dimension, which is expected to improve on the baseline evaluation done in this work.

Abstract: In this working note, we describe our approach to gastrointestinal disease and anatomical landmark classification for the Medico task at MediaEval 2017. We propose an inception-like CNN architecture and a fixed-crop data augmentation scheme for training and testing. The architecture is based on GoogLeNet and designed to keep the number of trainable parameters and its computational overhead small. Preliminary experiments show that the architecture is able to learn the classification problem from scratch using a tiny fraction of the provided training data only.

Abstract: The domain of minimally invasive surgery has recently attracted attention from the Multimedia community due to the fact that systematic video documentation is on the rise in this medical field. The vastly growing volumes of video archives demand for effective and efficient techniques to retrieve specific information from large video collections with visually very homogeneous content. One specific challenge in this context is to retrieve scenes showing similar surgical actions, i.e., similarity search. Although this task has a high and constantly growing relevance for surgeons and other health professionals, it has rarely been investigated in the literature so far for this particular domain. In this paper, we propose and evaluate a number of both static and dynamic content descriptors for this purpose. The former only take into account individual images, while the latter consider the motion within a scene. Our experimental results show that although static descriptors achieve the highest overall performance, dynamic descriptors are much more discriminative for certain classes of surgical actions. We conclude that the two approaches have complementary strengths and further research should investigate methods to combine them.

Abstract: The rapidly increasing volume of videos recorded in the course of endoscopic screenings and surgeries poses demanding challenges to video retrieval and browsing systems. Surgeons typically have to use standard video players to retrospectively review their procedures, which is an extremely cumbersome and time-consuming process. We present an HTML5-based video explorer that is specially tailored to this purpose and enables a time-efficient post-operative review of procedures. It incorporates various interactive browsing mechanisms as well as domain-specific content-based features based on previous research results. Preliminary interviews with surgeons indicate that this tool can considerably improve retrieval and browsing efficiency for users in the medical domain and allows surgeons to more easily and quickly revisit specific moments in recordings of their endoscopic surgeries.

Abstract: Although content-based retrieval methods achieved very good results for large-scale video collections in recent years, they still suffer from various deficiencies. On the other hand, plain human perception is a very powerful ability that still outperforms automatic methods in appropriate settings, but is very limited when it comes to large-scale data collections. In this paper, we propose to take the best from both worlds by combining an advanced content-based retrieval system featuring various query modalities with a straightforward mobile tool that is optimized for fast human perception in a sequential manner. In this collaborative system with multiple users, both subsystems benefit from each other: The results of issued queries are used to re-rank the video list on the tablet tool, which in turn notifies the retrieval tool about parts of the dataset that have already been inspected in detail and can be omitted in subsequent queries. The preliminary experiments show promising results in terms of search performance.

Abstract: Conversational services (e.g., Internet telephony) exhibit hard Quality of Service (QoS) requirements, such as low delay and jitter. Current IP-based solutions for conversational services use push-based data transfer only, since pull-based communication as envisaged in Named Data Networking (NDN) suffers from the two-way delay. Unfortunately, IP's addressing scheme requires additional services for contacting communication partners. NDN provides an inherent solution for this issue by using a location-independent naming scheme. Nevertheless, it currently does not provide a mechanism for push-based data transfer. In this paper, we investigate Persistent Interests as a solution for push-based communication. We improve and implement the idea of Persistent Interests, and study their applicability for conversational services in NDN. This is done by comparing different push- and pull-based approaches for Internet telephony.

Abstract: Persistent Interests (PIs) are a promising approach to introduce push-type traffic in Named Data Networking (NDN), in particular for conversational services such as voice and video calls. Forwarding decisions for PIs are crucial in NDN because they establish a long-lived path for the data flowing back toward the PI issuer. In the course of studying the use of PIs in NDN, we investigate adaptive PI forwarding and present a strategy combining regular NDN forwarding information and results from probing potential alternative paths through the network. Simulation results indicate that our adaptive PI forwarding approach is superior to the PI-adapted Best Route strategy when network conditions change due to link failures.

Abstract: When regarding physicians’ tremendously packed timetables, it comes as no surprise that they start managing even critical situations hastily in order to cope with the high demands laid out for them. Apart from treating patients’ conditions they as well are required to perform time-consuming administrative tasks, including post-surgery video analyses. Concerning documentation of minimally invasive surgeries (MIS), specifically endoscopy, such processes usually involve repeatedly perusing through lengthy, in the worst case uncut recordings – a redundant task that nowadays can be optimized by using readily available technology: we present a tool for annotating endoscopic video frames targeting a specific use case – endometriosis, i.e. the dislocation of uterine-like tissue.

Abstract: The nature of endoscopy as a type of minimally invasive surgery (MIS) requires surgeons to perform complex operations by merely inspecting a live camera feed. Inherently, a successful intervention depends upon ensuring proper working conditions, such as skillful camera handling, adequate lighting and removal of confounding factors, such as fluids or smoke. The latter is an undesirable byproduct of cauterizing tissue and not only constitutes a health hazard for the medical staff as well as the treated patients, it can also considerably obstruct the operating physician's field of view. Therefore, as a standard procedure the gaseous matter is evacuated by using specialized smoke suction systems that typically are activated manually whenever considered appropriate. We argue that image-based smoke detection can be employed to undertake such a decision, while as well being a useful indicator for relevant scenes in post-procedure analyses. This work represents a continued effort to previously conducted studies utilizing pre-trained convolutional neural networks (CNNs) and threshold-based saturation analysis. Specifically, we explore further methodologies for comparison and provide as well as evaluate a public dataset comprising over 100K smoke/non-smoke images extracted from the Cholec80 dataset, which is composed of 80 different cholecystectomy procedures. Having applied deep learning to merely 20K images of a custom dataset, we achieve Receiver Operating Characteristic (ROC) curves enclosing areas of over 0.98 for custom datasets and over 0.77 for the public dataset. Surprisingly, a fixed threshold for saturation-based histogram analysis still yields areas of over 0.78 and 0.75.

Abstract: The development and improper removal of smoke during minimally invasive surgery (MIS) can considerably impede a patient's treatment, while additionally entailing serious deleterious health effects. Hence, state-of-the-art surgical procedures employ smoke evacuation systems, which often still are activated manually by the medical staff or less commonly operate automatically utilizing industrial, highly-specialized and operating room (OR) approved sensors. As an alternate approach, video analysis can be used to take on said detection process -- a topic not yet much researched in aforementioned context. In order to advance in this sector, we propose utilizing an image-based smoke classification task on a pre-trained convolutional neural network (CNN). We provide a custom data set of over 30 000 laparoscopic smoke/non-smoke images, part of which served as training data for GoogLeNet-based [41] CNN models. To be able to compare our research for evaluation, we separately developed a non-CNN classifier based on observing the saturation channel of a sample picture in the HSV color space. While the deep learning approaches yield excellent results with Receiver Operating Characteristic (ROC) curves enclosing areas of over 0.98, the computationally much less costly analysis of an image's saturation histogram under certain circumstances can, surprisingly, as well be a good indicator for smoke with areas under the curves (AUCs) of around 0.92--0.97.

Abstract: An increasing number of surgeons promote video review of laparoscopic surgeries for detection of technical errors at an early stage as well as for training purposes. The reason behind is the fact that laparoscopic surgeries require specific psychomotor skills, which are difficult to learn and teach. The manual inspection of surgery video recordings is extremely cumbersome and time-consuming. Hence, there is a strong demand for automated video content analysis methods. In this work, we focus on retrieving surgical actions from video collections of gynecologic surgeries. We propose two novel dynamic content descriptors for similarity search and investigate a query-by-example approach to evaluate the descriptors on a manually annotated dataset consisting of 18 hours of video content. We compare several content descriptors including dynamic information of the segments as well as descriptors containing only spatial information of keyframes of the segments. The evaluation shows that our proposed dynamic content descriptors considering motion and spatial information from the segment achieve a better retrieval performance than static content descriptors ignoring temporal information of the segment at all. The proposed content descriptors in this work enable content-based video search for similar laparoscopic actions, which can be used to assist surgeons in evaluating laparoscopic surgical skills.

Abstract: We present an interface for interactive video browsing where users visually skim storyboard representations of the files in search for known items (known-item search tasks) and textually described subjects, objects, or events (ad-hoc search tasks). Individual segments of the video are represented as a color-sorted storyboard that can be addressed via a color-index. Our storyboard representation is optimized for quick visual inspections considering results from our ongoing research. In addition, a concept based-search is used to filter out parts of the storyboard containing the related concept(s), thus complementing the human-based visual inspection with a semantic, content-based annotation.

Abstract: We present an extension and evaluation of a novel interaction concept for video browsing on tablets. It can be argued that the best user experience for watching video on tablets can be achieved when the device is held in landscape orientation. Most mobile video players ignore this fact and make the interaction unnecessarily hard when the tablet is held with both hands. Naturally, in this hand posture only the thumbs are available for interaction. Our ThumbBrowser-interface takes this into account and combines it in its latest iteration with content analysis information as well as two different interaction methods. The interface was already introduced in a basic form in earlier work. In this paper we report on extensions that we applied and show first evaluation results in comparison to standard video players. We are able to show that our video browser is superior in terms of search accuracy and user satisfaction.

Abstract: In the domain of medical endoscopy an increasing number of surgeons nowadays store video recordings of their interventions in a huge video archive. Among some other purposes, the videos are used for post-hoc surgical quality assessment, since objective assessment of surgical procedures has been identified as essential component for improvement of surgical quality. Currently, such assessment is performed manually and for selected procedures only, since the amount of data and cumbersome interaction is very time-consuming. In the future, quality assessment should be carried out comprehensively and systematically by means of automated assessment algorithms. In this demo paper, we present a tool that supports human assessors in collecting manual annotations and therefore should help them to deal with the huge amount of visual data more efficiently. These annotations will be analyzed and used as training data in the future.

Abstract: Real-time entertainment services such as streaming audio- visual content deployed over the open, unmanaged Internet account now for more than 70% during peak periods. More and more such bandwidth hungry applications and services are proposed like immersive media services such as virtual reality and, specifically omnidirectional/360-degree videos. The adaptive streaming of omnidirectional video over HTTP imposes an important challenge on today’s video delivery infrastructures which calls for dedicated, thoroughly designed techniques for content generation, delivery, and consumption. This paper describes the usage of tiles — as specified within modern video codecs such HEVC/H.265 and VP9 — enabling bandwidth efficient adaptive streaming of omnidirectional video over HTTP and we define various streaming strategies. Therefore, the parameters and characteristics of a dataset for omnidirectional video are proposed and exemplary instanti- ated to evaluate various aspects of such an ecosystem, namely bitrate overhead, bandwidth requirements, and quality as- pects in terms of viewport PSNR. The results indicate bitrate savings from 40% (in a realistic scenario with recorded head movements from real users) up to 65% (in an ideal scenario with a centered/fixed viewport) and serve as a baseline and guidelines for advanced techniques including the outline of a research roadmap for the near future.