CStream: neighborhood bandwidth aggregation for better video streaming

Despite the popularity of watching videos online, challenges still remain in video streaming in many scenarios. Limited home broadband and mobile phone 3G bandwidths mean many users stream videos at compromised quality. To provide additional bandwidth for streaming, we propose CStream, a system that aggregates bandwidth from multiple cooperating users in a neighborhood environment for better video streaming. CStream exploits the fact that wireless devices have multiple network interfaces and connects cooperating users with a wireless ad-hoc network to aggregate their unused downlink Internet bandwidth. CStream dynamically generates a streaming plan to stream a single video using multiple connections, continuously adapting to changes in the neighborhood and variations in the available bandwidth. CStream is developed and evaluated on a test bed of computers, allowing for a detailed, controlled evaluation of performance. Analysis of the results shows a linear increase in throughput over single-connection streaming and improved video quality as the number of cooperating users in a neighborhood increase.     

一种基于用户播放行为序列的个性化视频推荐策略

本文针对在线视频服务网站的个性化推荐问题,提出了一种基于用户播放行为序列的个性化推荐策略.该策略通过深度神经网络词向量模型分析用户播放视频行为数据,将视频映射成等维度的特征向量,提取视频的语义特征.聚类用户播放历史视频的特征向量,建模用户兴趣分布矩阵.结合用户兴趣偏好和用户观看历史序列生成推荐列表.在大规模的视频服务系统中进行了离线实验,相比随机算法、基于物品的协同过滤和基于用户的协同过滤传统推荐策略,本方法在用户观看视频的Top-N推荐精确率方面平均分别获得22.3%、30.7%和934%的相对提升,在召回率指标上分别获得52.8%、41%和1065%的相对提升.进一步地与矩阵分解算法SVD++、基于双向LSTM模型和注意力机制的Bi-LSTM+Attention算法和基于用户行为序列的深度兴趣网络DIN比较,Top-N推荐精确率和召回率也得到了明显提升.该推荐策略不仅获得了较高的精确率和召回率,还尝试解决传统推荐面临大规模工业数据集时的数据要求严苛、数据稀疏和数据噪声等问题.     

一种基于P2P的短视频分享系统

现有在线短视频分享策略通常采用C/S架构,给视频服务器带来较大的带宽压力。为此,提出一种采用点对点方式的在线短视频分享系统IShare,该系统结合用户点播偏好和视频文件之间的社会网络特性实现视频分享。IShare主要包括基于点播兴趣的节点分簇和视频数据源节点的查找2个核心技术。实验结果表明,IShare具备较好的视频数据源节点查找能力,可降低视频服务器带宽资源消耗。     

Load balancing for video streaming services in hierarchical wireless networks

Wireless networks have focused on voice call services or wireless Internet access services. These days, the application service domain of wireless networks is rapidly expanding, and a wide variety of new services is emerging. Video streaming service is one of the most promising examples, evidenced by 3GPP’s MBMS (Multimedia Broadcast Multicast Service) and IMS (IP Multimedia Subsystem). In this paper, we consider the provision of video streaming services in hierarchical wireless networks with multiple layers of cells. We particularly focus on optimal load balancing among the cells, aiming at the minimization of frame drop ratio for given video streaming sessions. From this objective function, we derive the optimal load balancing condition. Load balancing is essentially the issue of which users are assigned to which cell, i.e., the user assignment problem. In our user assignment algorithm, we compute thresholds to divide users into groups according to the user characteristics, and map the user groups to proper cells. The optimal load balancing condition can be reached by adaptively adjusting the threshold at run time. This process does not require prior knowledge about the system status, such as the system capacity or user traffic requests, which warrants the practicality of the proposed scheme. Via simulations, we demonstrate that the proposed scheme achieves optimal load balancing in various realistic environments.     

Watching a Movie Alone yet Together: Understanding Reasons for Watching Danmaku Videos

By overlaying timeline-synchronized user comments on videos, Danmaku commenting creates a unique co-viewing experience of online videos. This study aims to understand the reasons for watching or not watching Danmaku videos. From a review of the literature and a pilot study, an initial pool of motivations and hindrances to Danmaku video viewing was gathered. Then, a survey involving 248 participants to identify the underlying factor structures of motivations and hindrances was conducted. Their influences on users’ attitude and behaviors with Danmaku videos were also examined. The results showed that people viewed Danmaku videos to obtain information, entertainment, and social connectedness. Introverted young men with high openness to new experience are more likely to view Danmaku videos. Infrequent viewers refused to watch Danmaku videos mainly because of the visual clutter that resulted from Danmaku comments.     

基于弹幕分析的在线直播平台用户理解

弹幕评论是网络直播平台与用户交互的主要方式之一,借助弹幕行为的分析可以更有效地实现对网络直播平台的用户理解.通过采集和利用3大热门直播平台(斗鱼、熊猫与战旗)的弹幕相关数据,本文以假设验证的方式从用户属性与用户行为两个角度对在线直播平台用户进行分析与理解,并建立基于用户行为特征时间序列的用户活跃模型对用户互动活跃度进行量化评估.研究表明,平台在线人数具有周期性变化的时间规律,观众地域具有沿海发达城市集中分布的空间取向,所提出的用户活跃模型能够对网络直播平台用户的行为活跃趋势做出合理的预测分析.     

Content-adaptive wireless streaming of instructional videos

Video streaming over wireless networks is becoming increasingly important for a variety of applications. To accommodate the dynamic change of wireless network bandwidths, Quality of Service (QoS) scalable video streams need to be provided. This paper presents a system of content-adaptive streaming of instructional (lecture) videos over wireless networks for E-learning applications. We first provide a real-time content analysis method to detect and extract content regions from instructional videos, then apply a “leaking-video-buffer” model to adjust QoS of video streams dynamically based on video content. In content-adaptive video streaming, an adaptive feedback control scheme is also developed to transmit properly compressed video streams to video clients not only based on network bandwidth, but also based on video content and the preferences of users. Finally, we demonstrate the scalability and content adaptiveness of the proposed video streaming system with experimental results on several instructional videos.     

Content-adaptive wireless streaming of instructional videos

Video streaming over wireless networks is becoming increasingly important for a variety of applications. To accommodate the dynamic change of wireless network bandwidths, Quality of Service (QoS) scalable video streams need to be provided. This paper presents a system of content-adaptive streaming of instructional (lecture) videos over wireless networks for E-learning applications. We first provide a real-time content analysis method to detect and extract content regions from instructional videos, then apply a “leaking-video-buffer” model to adjust QoS of video streams dynamically based on video content. In content-adaptive video streaming, an adaptive feedback control scheme is also developed to transmit properly compressed video streams to video clients not only based on network bandwidth, but also based on video content and the preferences of users. Finally, we demonstrate the scalability and content adaptiveness of the proposed video streaming system with experimental results on several instructional videos.     

Search-based composition, streaming and playback of video archive content

Locating content in existing video archives is both a time and bandwidth consuming process since users might have to download and manually watch large portions of superfluous videos. In this paper, we present two novel prototypes using an Internet based video composition and streaming system with a keyword-based search interface that collects, converts, analyses, indexes, and ranks video content. At user requests, the system can automatically sequence out portions of single videos or aggregate content from multiple videos to produce a single, personalized video stream on-the-fly.     

移动互联网视频传输优化的架构研究

随着3G移动互联网的快速发展,在手机等移动终端上看视频成为一种日常应用.但互联网上的大部分视频对于移动用户而言,其码率相对较大,而且移动用户的可用带宽受环境影响变化大,不稳定,影响了用户观看视频的体验.提出一种针对移动互联网的视频传输优化解决方案的系统架构,该架构从缓存和压缩两个方面入手,通过缓存缓解运营商的骨干网流量压力,通过视频压缩降低视频码率以满足用户的实际接入带宽.该架构既为运营商节约了带宽、降低了运营成本,同时也保证用户观看视频的连续性,提高用户体验.     

Affective video content analysis based on multimodal data fusion in heterogeneous networks

In heterogeneous networks, different modalities are coexisting. For example, video sources with certain lengths usually have abundant time-varying audiovisual data. From the users’ perspective, different video segments will trigger different kinds of emotions. In order to better interact with users in heterogeneous networks and improve their user experiences, affective video content analysis to predict users’ emotions is essential. Academically, users’ emotions can be evaluated by arousal and valence values, and fear degree, which provides an approach to quantize the prediction accuracy of the reaction of the audience and users towards videos. In this paper, we propose the multimodal data fusion method for integrating the visual and audio data in order to perform the affective video content analysis. Specifically, to align the visual and audio data, the temporal attention filters are proposed to obtain the time-span features of the entire video segments. Then, by using the two-branch network structure, matched visual and audio features are integrated in the common space. At last, the fused audiovisual feature is employed for the regression and classification subtasks in order to measure the emotional responses of users. Simulation results show that the proposed method can accurately predict the subjective feelings of users towards the video contents, which provides a way to predict users’ preferences and recommend videos according to their own demand.     

Video block device for user-friendly delivery in IaaS clouds

Peer-to-Peer (P2P) streaming technology based on various service requirements often remains on the elusive benefits of user-friendly video streaming in cloud computing environments. Cloud-assisted P2P media streaming gives an opportunity to enhance on-demand, dynamic and easily accessible videos. This paper outlines the fundamental video block device (VBD) method for user-friendly viewing patterns that inherits from user access patterns and provides efficient delivery using an enhanced BitTorrent protocol.     

Personalising live zooming using the ePlayer

An innovative Internet streaming video player, called ePlayer, oriented to live events, has been researched, developed and evaluated. ePlayer supports a personalised zoomable user interface which enables a new user view experience. The main novelty of the system is first that it is designed to optimise the zoomed video quality when viewing life events, via adaptation of the streamed video quality, across multi-video streams with a variable network quality of service. Second, it also personalises the live video zooming with respect to users’ zooming preferences, easing the user interaction needed for the zooming task. The experimental results indicate that the system is not only able to zoom effectively, but that it can also maintain the visual quality of the video at the same time. The ePlayer is also able to infer a user’s zooming preferences via dynamically clustering a user’s zooming regions of interest when viewing live sports video content.     

User-dependent interactive light field video streaming system

The sheer size and complex structure of light field (LF) videos bring new challenges to their compression and transmission. There have been numerous LF video compression algorithms reported in the literature to date. All of these algorithms compress and transmit all the views of an LF video. However, in some interactive or selective applications where users can choose the area of interest to be displayed, these algorithms generate a significant computational load and enormous data redundancies. In this paper, we propose an interactive LF video streaming system based on a user-dependent view selection scheme and an LF video coding method, which streams only the required data. Specifically, by predicting trajectories and using projection models, the viewing area of users in a limited consecutive number of time slots is firstly calculated, and then a user-dependent view selection method is proposed to determine the selected views of users for streaming. Finally, with the novel LF video sequence formed by only the selected sets of views, an adaptive coding method is presented for different LF video sequences based on users’ gestures. Experimental results illustrate that the proposed interactive LF video streaming system can achieve the best performance compared with other comparison methods.

     

MSC: a multi-version shared caching for multi-bitrate VoD services

Recently, many Video-on-Demand (VoD) service providers try to attract as many users as possible by offering multi-bitrate video streaming services with differentiated qualities. Many researches focus on video layered coding (e.g., scalable video coding, SVC). However, SVC is not widely used in VoD industry. Another solution, multi-version videos, can be classified into online transcoding and pre-stored multi-version videos. Online transcoding is a CPU-intensive and costly task, so it is not suitable for large-scale VoD applications. In this paper, we study how to improve caching efficiency based on pre-stored multi-version videos. We leverage the sharing probability among different versions of the same video and propose a multi-version shared caching (MSC) method to maximize the benefit of caching proxy. If the desired version is not in the cache while the higher neighbor version is in, MSC transmits the higher version streaming to user temporarily. In this case, MSC can make full use of the caching resources to improve the cache hit ratio and decrease users’ average waiting time. Simulation results show that MSC outperforms the others in the cache hit ratio and the average waiting time.     

A target advertisement system based on TV viewer’s profile reasoning

The traditional broadcasting services such as terrestrial, satellite and cable broadcasting have been unidirectional mass media regardless of TV viewer’s preferences. Recently rich media streaming has become possible via the broadband networks. Furthermore, since bidirectional communication is possible, personalcasting such as personalized streaming service has been emerging by taking into account the user’s preference on content genres, viewing times and actors/actresses etc. Accordingly personal media becomes an important means for content provision service in addition to the traditional broadcasting service as mass media. In this paper, we introduce a user profile reasoning method for TV viewers. The user profile reasoning is made in terms of genre preference and TV viewing times for TV viewer’s groups in different genders and ages. For user profiling reasoning, the TV viewing history data is used to train the proposed user profiling reasoning algorithm which allows for target advertisement for different age/gender groups. To show the effectiveness of our proposed user profile reasoning method, we present plenty of the experimental results by using real TV usage history.     

基于弹幕情感分析和聚类算法的视频用户群体分类

随着数字媒体等技术的发展,出现了弹幕系统这种新型的评论模式并逐渐流行。它能够使视频观众即时发布关于视频情节内容的评论,也可以帮助观众理解视频内容。弹幕文本数据的产生,为短文本处理和实时数据处理提供了新的素材。研究弹幕数据的特点和其表达的情感,可以帮助我们更好地理解视频情节;研究弹幕内容之间的相似度进而分析用户之间的关联关系,不仅能够深入了解弹幕用户的特点、发掘不同视频之间的潜在联系,而且可以为视频制作时受众群体的选择提供更为准确的解决方案。首先将弹幕文本数据进行收集和预处理,然后计算这些文本的情感值。针对弹幕文本口语化的特点,建立了网络弹幕常用词词典。通过改进传统的k-means聚类算法,对所有发表弹幕的用户进行基于情感值的分类。这样的分类可以帮助我们了解观看特定类型视频的观众在情感上的异同点。     

Discrimination of media moments and media intervals: sticker-based watch-and-comment annotation

In this paper we discuss the problem of how to discriminate moments of interest on videos or live broadcast shows. The primary contribution is a system which allows users to personalize their programs with previously created media stickers??pieces of content that may be temporarily attached to the original video. We present the system??s architecture and implementation, which offer users operators to transparently annotate videos while watching them. We offered a soccer fan the opportunity to add stickers to the video while watching a live match: the user reported both enjoying and being comfortable using the stickers during the match??relevant results even though the experience was not fully representative.     

Collaborative Interaction for Videos on Mobile Devices Based on Sketch Gestures

With the rapid progress of the network and mobile techniques, mobile devices such as mobile phones and portable devices, have become one of the most important parts in common life.Efficient techniques for watching, navigating and sharing videos on mobile devices collaboratively are appealing in mobile environment.In this paper, we propose a novel approach supporting efficiently collaborative operations on videos with sketch gestures.Furthermore, effective collaborative annotation and navigation operations are given to interact with videos on mobile devices for facilitating users’ communication based on mobile devices’ characteristics.Gesture operation and collaborative interaction architecture are given and improved during the interactive process.Finally, a user study is conducted showing that the effectivity and collaborative accessibility of video exploration is improved.     

User demographics prediction based on mobile data

Demographics prediction is an important component of user profile modeling. The accurate prediction of users’ demographics can help promote many applications, ranging from web search, personalization to behavior targeting. In this paper, we focus on how to predict users’ demographics, including “gender”, “job type”, “marital status”, “age” and “number of family members”, based on mobile data, such as users’ usage logs, physical activities and environmental contexts. The core idea is to build a supervised learning framework, where each user is represented as a feature vector and users’ demographics are considered as prediction targets. The most important component is to construct features from raw data and then supervised learning models can be applied. We propose a feature construction framework, CFC (contextual feature construction), where each feature is defined as the conditional probability of one user activity under the given contexts. Consequently, besides employing standard supervised learning models, we propose a regularized multi-task learning framework to model different kinds of demographics predictions collectively. We also propose a cost-sensitive classification framework for regression tasks, in order to benefit from the existing dimension reduction methods. Finally, due to the limited training instances, we employ ensemble to avoid overfitting. The experimental results show that the framework achieves classification accuracies on “gender”, “job” and “marital status” as high as 96%, 83% and 86%, respectively, and achieves Root Mean Square Error (RMSE) on “age” and “number of family members” as low as 0.69 and 0.66 respectively, under the leave-one-out evaluation.