Search

The video posting service is a web service that is familiar to many people and is expected to be in demand in the future. In the video posting site Nico Nico Douga, we proposed an interface in which user comments are superimposed on the video and played on the screen. In this study, we examined a new video posting service that allows users to post their comments not as text, but as audio, and verified the possibility of such a service. As a result, through the prototyping of the system, it became clear that it has the potential to become an unprecedented new video posting service. On the other hand, the proposed system has some problems, such as the difficulty of cancelling voice comments afterwards. In order to solve these problems, we need to devise a better operation method and improve the system. In the future, we will improve these issues through user evaluation.

As smart environments become laden with more and more sensors, there has been a need to develop systems that could derive useful information from these sensors and make the smart environments smarter. Complex Event Processing (CEP) has emerged as a popular strategy to identify crucial events from sensor data. However, the existing CEP strategies overlook the relationship with other sensors in the spatial vicinity and understate the temporal variation of sensor data. In this paper, we develop RACER, which is an end-to-end complex event processing system that takes into consideration both the spatial location of the sensor in observation and the varying impact of temporal changes in the sensor data. Experiments performed for a duration of five months over both collected and live streaming data shows that RACER fares well compared to the other state-of-the-art approaches.

The use of online food delivery (OFD) services has skyrocketed despite lingering concerns about food safety during the COVID-19 pandemic. The pandemic has impacted most business sectors, including local farms and ranches whose channels to sell through local food markets have shut down. With the demand for food delivery is increasing while the local food market at risk, we saw an opportunity to connect these two complementary businesses through “live-streaming farms.” We conducted interviews to investigate the current challenges of using OFD services, designed and evaluated the concept of live-streaming farms in OFD services to promote sustainable local food systems. Our findings indicate that offering extra food-related information through live streaming could potentially reduce primary concerns about food safety for OFD users and build local food networks. We conclude by discussing the design considerations of live-streaming farms that can improve user experience with OFD and support the local food systems.

Live-streaming platforms have recently gained significant popularity by attracting an increasing number of young users and have become a very promising form of online shopping. Similar to the traditional online shopping platforms such as Taobao, live-streaming platforms also suffer from online malicious fraudulent behaviors where many transactions are not genuine. The existing anti-fraud models proposed to recognize fraudulent transactions on traditional online shopping platforms are inapplicable on live-streaming platforms. This is mainly because live-streaming platforms are characterized by a unique type of heterogeneous live-streaming networks where multiple heterogeneous types of nodes such as users, live-streamers, and products are connected with multiple different types of edges associated with edge features. In this paper, we propose a new approach based on a heterogeneous graph neural network for LIve-streaming Fraud dEtection (called LIFE). LIFE designs an innovative heterogeneous graph learning model that fully utilizes various heterogeneous information of shopping transactions, users, streamers, and items from a given live-streaming platform. Moreover, a label propagation algorithm is employed within our LIFE framework to handle the limited number of labeled fraudulent transactions for model training. Extensive experimental results on a large-scale Taobao live-streaming platform demonstrate that the proposed method is superior to the baseline models in terms of fraud detection effectiveness on live-streaming platforms. Furthermore, we conduct a case study to show that the proposed method is able to effectively detect fraud communities for live-streaming e-commerce platforms.

It is highly challenging to simultaneously achieve high-rate and low-latency in live video streaming. Chunk-based streaming and playback speed adaptation are two promising new trends to achieve high user Quality-of-Experience (QoE). To thoroughly understand their potentials, we develop a detailed chunk-level dynamic model that characterizes how video rate and playback speed jointly control the evolution of a live streaming session. Leveraging on the model, we first study the optimal joint video rate-playback speed adaptation as a non-linear optimal control problem. We further develop model-free joint adaptation strategies using deep reinforcement learning. Through extensive experiments, we demonstrate that our proposed joint adaptation algorithms significantly outperform rate-only adaptation algorithms and the recently proposed low-latency video streaming algorithms that separately adapt video rate and playback speed without joint optimization. In a wide-range of network conditions, the model-based and model-free algorithms can achieve close-to-optimal trade-offs tailored for users with different QoE preferences.

Over the past few years, the ubiquitous usage of internet to broadcast information worldwide, has proved to be one of the best methods in making people aware about their surroundings. This has also led towards storage of vast amount of data in user interactive websites. Several news channels apart from live streaming, are using internet in such ways to convey their information. And these methods have not only benefited people to acquire regular updates but have also impacted their lives in many ways during the world awakening pandemic like COVID-19. Currently, in this pandemic situation, many leaders including federal and state government officials have shown great concern towards society by providing statements such as preventive measures, which were regularly recorded and elaborated in the form of stories by the news channels in their own websites. But, in general, the type of statement given by a leader will always affect people's opinion and behavior, which also happened during this pandemic. There were variations in number of cases of COVID-19 based upon a region leader's statements as per timeline and the people's modulating lifestyle there. This kind of relationship has motivated us towards analyzing the COVID-19 data based on sentiment and emotion involved in leader's statements, giving rise to an idea of web scraping to obtain their statements and stories from news channels. The main aim of this paper is to enlighten about different techniques and libraries used for web scraping, some challenges while designing a web scraper and finally arrive at an efficient methodology using them to extract news headlines and stories and create a new dataset that can be used in sentiment and emotion analysis. In future work, we will present the results of our experiments with the downloaded datasets.

Twitch is a live streaming platform that was originally designed for gamers to share their gaming experiences. The viewer culture on Twitch supports a lively exchange between streamers and viewers and also among viewers themselves. While the potential of Twitch for game tutorials and instructions has been explored, its potential for university lectures still appears to be undiscovered. In this paper, we discuss the potential of Twitch for teaching. We ran two different courses on Twitch, one for CS beginners, the other a game development course for students in a master's program. Evaluations with students show Twitch benefits, including increased engagement and interaction in chat and the ability to watch videos directly afterward. On the other hand, we also discovered several challenges, such as in-chat spam by bots. In this paper, we would like to discuss initial experiences from different perspectives and discuss best practices. Our aim in doing so, is to help other teachers decide whether Twitch might be a suitable online learning platform for their course.

There are two main factors that determine the viewer experience during the live streaming of sports content: latency and stalls. Latency should be low and stalls should not occur. Yet, these two factors work against each other and it is not trivial to strike the best trade-off between them. One of the best tools we have today to manage this trade-off is the adaptive playback speed control. This tool allows the streaming client to slow down the playback when there is a risk of stalling and increase the playback when there is no risk of stalling but the live latency is higher than desired. While adaptive playback generally works well, the artifacts due to the changes in the playback speed should preferably be unnoticeable to the viewers. However, this mostly depends on the portion of the audio/video content subject to the playback speed change. In this paper, we advance the state-of-the-art by developing a content-aware playback speed control (CAPSC) algorithm and demonstrate a number of examples showing its significance. We make the running code available and provide a demo page hoping that it will be a useful tool for the developers and content providers.

Reducing end-to-end streaming latency is critical to HTTP-based live video streaming. There are currently two technologies in this domain: Low-Latency HTTP Live Streaming (LL-HLS) and Low-Latency Dynamic Adaptive Streaming over HTTP (LL-DASH). Many players support LL-HLS and/or LL-DASH protocols, including Apple's AVPlayer, Shaka player, HLS.js Dash.js, and others. This paper is dedicated to the analysis of the performance of low-latency players and streaming protocols. The evaluation is based on a series of live streaming experiments, repeated using identical video content, encoders, encoding profiles, and network conditions, emulated by using traces of real-world networks. Several performance metrics, such as average stream bitrate, the amounts of downloaded media data, streaming latency, as well as buffering and stream switching statistics are captured and reported in our experiments. These results are subsequently used to describe the observed differences in the performance of LL-HLS and LL-DASH-based players.

Rules and norms are critical to community governance. Live streaming communities like Twitch consist of thousands of micro-communities called channels. We conducted two studies to understand the micro-community rules. Study one suggests that Twitch users perceive that both rules transparency and communication frequency matter to channel vibe and frequency of harassment. Study two finds that the most popular channels have no channel or chat rules; among these having rules, rules encouraged by streamers are prominent. We explain why this may happen and how this contributes to community moderation and future research.

Live streaming is a form of interactive media that potentially makes streamers more vulnerable to harassment due to the unique attributes of the technology that facilitates enhanced information sharing via video and audio. In this study, we document the harassment experiences of 25 live streamers on Twitch from underrepresented groups including women and/or LGBTQ streamers and investigate how they handle and prevent adversity. In particular, live streaming enables streamers to self-moderate their communities, so we delve into the methods of how they manage their communities from both a social and technical perspective. We found that technology can cover the basics for handling negativity, but much emotional and relational work is invested in moderation, community maintenance, and self-care.

Cyberbullying is widespread in online communities. The ability to protect users and platforms from the dangers of cyberbullying has become essential. In this work, we conducted a qualitative study of Twitch community to explore the design elements that can influence cyberbullying behavior and the reaction to it on a live-streaming platform. Our study involved interviewing 11 Twitch streamers and six months of Reddit posts from a Twitch sub-reddit community about Cyberbullying. The results of the study illustrate a number of design themes unique to live streaming platform that can promote or inhibit cyberbullying, including most importantly the information asymmetry between broadcaster and viewers. Based on our analysis results, we propose design guidelines for the live streaming platform to better support their users in intervening cases of cyberbullying.

With the development of the Internet, the function of mobile e-commerce webcast on e-commerce platforms has begun to develop. At the beginning of this year, because of COVID-19, people were more shopping at home by watching webcast, which promoted the further development of mobile e-commerce webcast. Mobile e-commerce webcast refers to the display of abundant products on mobile e-commerce platform merchants through live streaming media. In this paper, combined with the environmental characteristics of mobile e-commerce webcast, based on the Stimulus-Organism-Response (S-O-R) theoretical model, a model of the influencing factors of users’ online shopping behavior intentions is constructed. Taking users of China's major mobile e-commerce webcast platforms as the research object, through structural equation model analysis, the factors affecting users' willingness to use online shopping under mobile e-commerce webcast are explored.

We propose a study on the effects of live streaming on an educational game’s learning outcomes. The COVID-19 pandemic has strengthened the call for interactive online learning experiences. There is a growing body of literature examining learning on platforms such as Twitch, and studies have shown that enhancing in-game performance is possible from viewing a stream. However, little work has explored whether learning from live streaming educational games, where in-game performance relates to educational outcomes outside of the game context, is possible. We share the details of our proposed study, in which an educational game called Angle Jungle will streamed to participants, and learning gains will be compared to three non-live streamed conditions. By executing this study, we can understand the benefits and shortcomings of current live streaming interfaces in supporting educational games, paving the way for the design of novel and interactive learning experiences built for live streaming platforms.

Live streaming is a rapidly growing industry, with millions of content creators using platforms like Twitch to share games, art, and other activities. However, with this rise in popularity, most streamers often fail to attract viewers and grow their platforms. Analytic tools—which have shown success in other business and learning contexts—may be one potential solution, but their use in streaming settings remains unexplored. In this study, we focused on game streaming and interviewed 18 game streamers on Twitch and Mixer about their information needs and current use of tools, supplemented by explorations into their Discord communities. We find that streamers have a range of content, marketing, and community information needs, many of which are not being met by available tools. We conclude with design implications for developing more streamer-centered analytics for video game streamers.

Live streaming has become increasingly popular, with most streamers presenting their real-life appearance. However, Virtual YouTubers (VTubers), virtual 2D or 3D avatars that are voiced by humans, are emerging as live streamers and attracting a growing viewership in East Asia. Although prior research has found that many viewers seek real-life interpersonal interactions with real-person streamers, it is currently unknown what makes VTuber live streams engaging or how they are perceived differently than real-person streamers. We conducted an interview study to understand how viewers engage with VTubers and perceive the identities of the voice actors behind the avatars (i.e., Nakanohito). The data revealed that Virtual avatars bring unique performative opportunities which result in different viewer expectations and interpretations of VTuber behavior. Viewers intentionally upheld the disembodiment of VTuber avatars from their voice actors. We uncover the nuances in viewer perceptions and attitudes and further discuss the implications of VTuber practices to the understanding of live streaming in general.

The COVID-19 global pandemic and resulted lockdown policies have forced education in nearly every country to switch from a traditional co-located paradigm to a pure online 'distance learning from home' paradigm. Lying in the center of this learning paradigm shift is the emergence and wide adoption of distance communication tools and live streaming platforms for education. Here, we present a mixed-methods study on live streaming based education experience during the COVID-19 pandemic. We focus our analysis on Chinese higher education, carried out semi-structured interviews on 30 students, and 7 instructors from diverse colleges and disciplines, meanwhile launched a large-scale survey covering 6291 students and 1160 instructors in one leading Chinese university. Our study not only reveals important design guidelines and insights to better support current remote learning experience during the pandemic, but also provides valuable implications towards constructing future collaborative education supporting systems and experience after pandemic., Comment: Zhilong Chen and Hancheng Cao contribute equally to this work; Accepted to CHI 2021

Nowadays, the industry of network live broadcasting platform is developing rapidly, which attracts people's attention. And in the network live broadcast, live with goods or network anchor recommended goods and other behavior has been common. Especially after the epidemic, through the network live marketing, has become an important means to promote economic recovery. However, the level of product quality in live network broadcasting is not uniform, and the problem that consumers' rights and interests cannot be protected is gradually revealed. Therefore, this paper discusses the role of e-commerce law based on big data on live network. In the discussion, this paper first analyzes the e-commerce law to clarify the applicability of the e-commerce law in the live network; secondly, through the investigation of the network anchor and fans, analyzes the business behavior in the network live broadcast; finally, analyzes the role of the e-commerce law based on big data on the network live broadcast. The results show that the e-commerce law based on big data has a good regulatory effect on webcast, which can promote the healthy development of webcast.

Transcoding is essential for live video streaming systems, but video streams in many channels may not be transcoded due to the processing capacity limitations of transcoding servers with multiple CPU nodes. To address this, we propose an algorithm that determines the transcodeable bitrate versions for each channel and allocates the transcoding tasks, by taking into account video quality, popularity, and workload balancing. Simulation results show that it can improve popularity-weighted video quality by up to 10.7% over other popularity-based alternatives.

As more viewers become interested in watching authentic personalities as opposed to artificial, crafted performances, Twitch streamers have begun to discuss personal issues such as mental health to form a closer bond with their community. This paper seeks to further explore how a live streamer's choice in game affects their viewers' perception of their stream's content. We interviewed 24 Twitch viewers and found that the pace of the game a streamer chooses to stream is a key factor in what viewers expect to see during the stream. These expectations at least partially determine whether viewers want to hear conversations about mental health.

In this project, we explore the intersection of improvisational theater and Twitch live streaming. Improv offers social affordances to live streams, while streaming offers improv performers a way to both broaden their reach to non-copresent audiences and create new types of participatory improv games. Based on insights from both domains, we present StAGE, a live streaming system that both communicates performative information otherwise lost when streaming live theater and accommodates participatory roles for performers, hosts, and audiences. Using our system, we created and playtested four improv game prototypes exploring a range of interactions. We show that StAGE affords novel interactions both as a streaming experience and as improvisational theater. We propose that the social transfer of roles and information, particularly by hosts, is a critical aspect of the design. We conclude by discussing newly discovered challenges and potential for the future of the intersection of these domains.

Live streaming is a new type of social media that uses streaming media technologies for broadcasting video in real time. The rise of live streaming services has changed consumer behaviors and brought up a new business model. Usually, live streamer is able to receive a constant income from commissions of user subscriptions and earn extra money from product sales and user donations. Prior studies focused on how to acquire viewers; while they failed to explore how to transform the viewers watching for free into valuable consumers capable of purchasing, subscribing, and even donating to streamers. This study explored the feasibility of finding potential subscribers. A predictive model for potential subscriber is built by the machine learning approach based on users' personal information and their online interaction with streamers. The preliminary result shows the predictive model constructed by the random-forest technique using daily interaction data can better predict the potential subscribers.

Live streaming is an increasingly popular communication medium that allows real-time interaction among a broadcaster and an audience of any size. Using archived YouTube live video transcripts and associated live chat messages, we find evidence for emotional contagion in live streams: sentiment in live video oral transcripts and viewers? text chat is associated with the sentiment in subsequent viewers? comments. This relationship is stronger between viewers? chat messages and the subsequent chat than between the oral messages in the video and the subsequent chat. However, in some types of live streams, negative sentiment in the live video is followed by less negative chat. We conclude with a discussion of future research and potential uses of the dataset.

Nowadays, people spend dramatically more time on watching videos through different devices. The advanced hardware technology and network allow for the increasing demands of users viewing experience. Thus, enhancing the Quality of Experience of end-users in advanced multimedia is the ultimate goal of service providers, as good services would attract more consumers. Quality assessment is thus important. The first workshop on "Quality of Experience (QoE) in visual multimedia applications" (QoEVMA'20) focuses on the QoE assessment of any visual multimedia applications both subjectively and objectively. The topics include 1)QoE assessment on different visual multimedia applications, including VoD for movies, dramas, variety shows, UGC on social networks, live streaming videos for gaming/shopping/social, etc. 2)QoE assessment for different video formats in multimedia services, including 2D, stereoscopic 3D, High Dynamic Range (HDR), Augmented Reality (AR), Virtual Reality (VR), 360, Free-Viewpoint Video(FVV), etc. 3)Key performance indicators (KPI) analysis for QoE. This summary gives a brief overview of the workshop, which took place at October 16, 2020 in Seattle (U.S.), as a half-day workshop.

Is it possible to predict moment-to-moment gameplay engagement based solely on game telemetry? Can we reveal engaging moments of gameplay by observing the way the viewers of the game behave? To address these questions in this paper, we reframe the way gameplay engagement is defined and we view it, instead, through the eyes of a game's live audience. We build prediction models for viewers' engagement based on data collected from the popular battle royale game PlayerUnknown's Battlegrounds as obtained from the Twitch streaming service. In particular, we collect viewers' chat logs and in-game telemetry data from several hundred matches of five popular streamers (containing over 100,000 game events) and machine learn the mapping between gameplay and viewer chat frequency during play, using small neural network architectures. Our key findings showcase that engagement models trained solely on 40 gameplay features can reach accuracies of up to 80% on average and 84% at best. Our models are scalable and generalisable as they perform equally well within- and across-streamers, as well as across streamer play styles., Comment: Version accepted for the Conference on the Foundations of Digital Games 2020 - Malta

Live video broadcasts to particular communities or targeted audiences many-at-a-times indulge a cluster of localities from where the end-users are interested in participating. In this paper, we leverage this idea to develop a system called Federated Live Streaming over DASH (FLiDASH) which forms end-users coalitions based on the locality of the network connectivity (like a set of end-users behind a common core-network service gateway). In FLiDASH, the coalition members stream the live data collectively based on a federated adaptive bitrate streaming mechanism where the coalition as a whole decides the optimal bitrate for video streaming and distributes the download overhead among its members. We have thoroughly evaluated FLiDASH in an emulated setup, and observe a 40% improvement in the live streaming QoE with a 20% reduction in the network traffic usage.

The outbreak of covid-19 since December 2019, as a short-term accelerator, has helped web celebrity economic e-commerce live broadcast industry enter the public eye. This paper first takes the supply and demand theory framework as the underlying logic to discuss the current situation. From the perspective of policy support, economic support, social development and technological progress, this essay expounds the content of demand side. On the supply side, the historical evolution of web celebrity economic live broadcast industry is detailed. In addition, the author analyzes the industrial chain of the industry at length, and calculates the benefit distribution mechanism of different platforms. Finally, based on the above analysis, this article predicts the future development trend of this industry and puts forward possible risk factors, hoping to come up with feasible suggestions for the development of web celebrity live e-commerce economy.

Live video accounts for a significant volume of today's Internet video. Despite a large number of efforts to enhance user quality of experience (QoE) both at the ingest and distribution side of live video, the fundamental limitations are that streamer's upstream bandwidth and computational capacity limit the quality of experience of thousands of viewers. To overcome this limitation, we design LiveNAS, a new live video ingest framework that enhances the origin stream's quality by leveraging computation at ingest servers. Our ingest server applies neural super-resolution on the original stream, while imposing minimal overhead on ingest clients. LiveNAS employs online learning to maximize the quality gain and dynamically adjusts the resource use to the real-time quality improvement. LiveNAS delivers high-quality live streams up to 4K resolution, outperforming WebRTC by 1.96 dB on average in Peak-Signal-to-Noise-Ratio on real video streams and network traces, which leads to 12%-69% QoE improvement for live stream viewers.

Recently, live streaming has been becoming a new service provided by social commerce platforms, enabling small sellers to sell products online. Compared with traditional electronic commerce, the immediately interpersonal interaction is a vital feature of live streaming shopping. Based on Parasocial Interaction (PSI) Theory, this study aimed to empirically explore the influence of interpersonal interaction factors on viewers' PSI perception towards streamers. The relationship between viewers' PSI perception towards streamers and their shopping intention was also investigated. This study used Facebook Live as a research field. The results suggest that the viewers would form a positive perception of expertise, familiarity, and similarity to live streamers while watching their live streaming selling channel. Meanwhile, they could perceive the social presence of other viewers from the chat room, which is a significant function of live streaming services. Among these interpersonal interaction factors, familiarity is the primary factor increasing the viewers' PSI perception towards live streamers, followed by social presence and similarity. The results also reveal that the viewers' PSI perception towards live streamers could further drive their shopping intention. Finally, this study provides several practical suggestions for live streamers to sell products on live streaming platforms.

Live streaming is a unique medium that merges different layers of communication by facilitating individual, group, and mass communication simultaneously. Streamers who broadcast themselves on live streaming platforms such as Twitch are their own media entity and have the challenge of having to manage interactions with many different types of online audiences beyond the translucent platform interfaces. Through qualitative interviews with 25 Twitch streamers, in this paper we share streamers’ practices of discovering audience composition, categorizing audience groups, and developing appropriate mechanisms to interact with them despite geographical, technological, and temporal limitations. We discuss streamers’ appropriation of real-time signals provided by these platforms as sources of information, and their dependence on both technology and voluntary human labor to scale their media entity. We conclude with design recommendations for streaming platforms to provide streamer-centric tools for audience management, especially for knowledge discovery and growth management. .

As video calls and personal broadcasting become popular, the demand for mobile live streaming over cellular uplink channels is growing fast. However, current live streaming solutions are known to suffer from frequent uplink throughput fluctuations causing unnecessary video stalls and quality drops. As a remedy to this problem, we propose PERCEIVE, a deep learning-based uplink throughput prediction framework. PERCEIVE exploits a 2-stage LSTM (Long Short Term Memory) design and makes throughput predictions for the next 100ms. Our extensive evaluations show that PERCEIVE, trained with LTE network traces from three major operators in the U.S., achieves high accuracy in the uplink throughput prediction with only 7.67% mean absolute error and outperforms existing prediction techniques. We integrate PERCEIVE with WebRTC, a popular video streaming platform from Google, as a rate adaptation module. Our implementation on the Android phone demonstrates that it can improve PSNR by up to 6dB (4x) over the default WebRTC while providing less streaming latency.

Given the central role of webcams in monitoring physical surroundings, it behooves the research community to understand the characteristics of webcams' distribution and their privacy/security implications. In this paper, we conduct the first systematic study on live webcams from both aggregation sites and individual webcams (webpages/IP hosts). We propose a series of efficient, automated techniques for detecting and fingerprinting live webcams. In particular, we leverage distributed algorithms to detect aggregation sites and generate webcam fingerprints by utilizing the Graphical User Interface (GUI) of the built-in web server of a device. Overall, we observe 0.85 million webpages from aggregation sites hosting live webcams and 2.2 million live webcams in the public IPv4 space. Our study reveals that aggregation sites have a typical long-tail distribution in hosting live streams (5.8% of sites contain 90.44% of live streaming contents), and 85.4% of aggregation websites scrape webcams from others. Further, we observe that (1) 277,239 webcams from aggregation sites and IP hosts (11.7%) directly expose live streams to the public, (2) aggregation sites expose 187,897 geolocation names and more detailed 23,083 longitude/latitude pairs of webcams, (3) the default usernames and passwords of 38,942 webcams are visible on aggregation sites in plaintext, and (4) 1,237 webcams are detected as having been compromised to conduct malicious behaviors.

HTTP Adaptive Streaming (HAS) technologies such as MPEG DASH are now used extensively to deliver television services to large numbers of viewers. In HAS, the client requests segments of content using HTTP, with an ABR algorithm selecting the quality at which to request each segment to trade-off video quality with the avoidance of stalling. This introduces significant end to end latency compared to traditional broadcast, due to the the client requiring a large enough buffer for the ABR algorithm to react to changes in network conditions in a timely manner. The recently standardised Common Media Application Format (CMAF) has helped address the issue of latency by defining segments as composed of independently transferable chunks. In this paper, we describe a simulation model we have developed to evaluate the performance of four popular ABR algorithms using DASH and CMAF in various low latency live streaming scenarios. Realistic network conditions are used for the evaluation, which are based on throughput data taken from the CDN logs of a commercial live TV service. We quantify the performance of the ABR algorithms using a selection of QoE metrics, and show that CMAF can significantly improve ABR performance in low delay scenarios.

Live streaming remains a challenge in the adaptive streaming space due to the stringent requirements for not just quality and rebuffering, but also latency. Many solutions have been proposed to tackle streaming in general, but only few have looked into better catering to the more challenging low-latency live streaming scenarios. In this paper, we re-visit and extend several important components (collectively called Low-on-Latency, LoL) in adaptive streaming systems to enhance the low-latency performance. LoL includes bitrate adaptation (both heuristic and learning-based), playback control and throughput measurement modules.

Live streaming visual art such as drawing or using design software is gaining popularity. An important aspect of live streams is the direct and real-time communication between streamers and viewers. However, currently available text-based interaction limits the expressiveness of viewers as well as streamers, especially when they refer to specific moments or objects in the stream. To investigate the feasibility of using snapshots of streamed content as a way to enhance streamer-viewer interaction, we introduce Snapstream, a system that allows users to take snapshots of the live stream, annotate them, and share the annotated snapshots in the chat. Streamers can also verbally reference a specific snapshot during streaming to respond to viewers' questions or comments. Results from live deployments show that participants communicate more expressively and clearly with increased engagement using Snapstream. Participants used snapshots to reference part of the artwork, give suggestions on it, make fun images or memes, and log intermediate milestones. Our findings suggest that visual interaction enables richer experiences in live streaming.

Many artists broadcast their creative process through live streaming platforms like Twitch and YouTube, and people often watch archives of these broadcasts later for learning and inspiration. Unfortunately, because live stream videos are often multiple hours long and hard to skim and browse, few can leverage the wealth of knowledge hidden in these archives. We present an approach for automatic temporal segmentation of creative live stream videos. Using an audio transcript and a log of software usage, the system segments the video into sections that the artist can optionally label with meaningful titles. We evaluate this approach by gathering feedback from expert streamers and comparing automatic segmentations to those made by viewers. We find that, while there is no one "correct" way to segment a live stream, our automatic method performs similarly to viewers, and streamers find it useful for navigating their streams after making slight adjustments and adding section titles.

Live streaming platforms not only provide live videos but also allow social interactions between viewers via real-time chatting. However, none of existing research has studied the social impact for recommending live streams. In this work, we formulate a new personalized recommendation problem by factoring in both video and social contents (chats). Accordingly, we 1) design a new attention network ANSWER to identify viewers’ attention on video and social contents, and 2) rank the channels based on the attentive features. We collect a real dataset from Twitch for evaluation. The experimental results manifest that ANSWER outperforms baselines by at least 26.6% in terms of NDCG@5.

Live streaming has become a ubiquitous channel for people to learn new happenings. Although live streaming videos generally attract a large audience of watchers, their contents are long and contain relatively unexciting stretches of knowledge transmission. This observation has prompted artificial intelligence researchers to establish advanced models that automatically extract highlights from live streaming videos. Most streaming highlight extraction research has been based on visual analysis of video frames, and seldom have studies considered the messages posted by the audiences. In this paper, we propose a deep learning model that examines the messages posted by streaming audiences. The video segments whose messages reveal audience excitement are extracted to compose the highlights of a streaming video. We evaluate our model in terms of multiple Twitch streaming channels. The precision of our highlight extraction model is 51.3% and is superior to several baseline methods.

This paper investigates whether subscribers (paying viewers) can be identified by their chat messages in live streaming services, Twitch. Collecting data of twenty streamers on Twitch, this study conducts two studies: (i) presenting sentimental differences of viewers' chat messages between unsubscribers (free viewers) and subscribers (paying viewers), and (ii) employing a set of machine learning-based classifiers which attempt to accurately identify subscribers. The results show that the machine learning-based classifiers with sentimental features provide the potentiality to identify paying viewers, suggesting their utility as a research toolkit.

Harassment is an issue in online communities with the live streaming platform Twitch being no exception. In this study, we surveyed 375 Twitch users in person at TwitchCon, asking them about who should be responsible for deciding what should be allowed and what strategies they perceived to be effective in handling harassment. We found that users thought that streamers should be most responsible for enforcing rules and that either blocking bad actors, ignoring them, or trying to educate them were the most effective strategies.

Charity streaming is a novel and increasingly popular form of fundraising where content creators stream content during a fixed period of time to raise money and awareness. Many charity streams involve people playing games for a prolonged period of time. In this study, we interviewed charity organizations about what this new form of fundraising means for them and how it is different from traditional fundraising efforts.

Live quiz shows that make use of mobile applications embed aspects of both games and live streaming. To understand users' motivations and experiences related to these mobile quiz shows, we used a mixed-methods approach that includes interviews (N = 16) and a survey (N = 296). We conducted a thematic analysis of interviews to identify seven motivations, five of which were verified through a factor analysis of survey data: interactivity, offline integration, achievement, ease of enjoyment, and financial incentives. Moreover, we examined how those motivations influenced users' perceptions of the apps and their in-app behavior patterns. The motivations of offline integration and financial incentives positively affected users' frequency of app usage, their recommendation of the app, and their habitual usage. This result implies that, to drive persistent usage, it is important to foster social integration and to properly balance both the average chance of winning and the amount of distributed prize money. Furthermore, our results suggest new directions for using design implications to improve user engagement.

Digital patronage is the act of delivering recurring direct support to content creators online. In this paper, we define digital patronage and examine why patrons engage in this behavior on the live streaming platform Twitch. Our mixed method research illustrates patrons' motivations, how patronage motivations differ from that of donations, and the motivational factors that are associated with higher levels of patronage. We discuss how results extend understanding of patronage in the context of social support theory and provide design implications for digital patronage platforms.

We demonstrate a video-to-animation translator, which can transform real-world video into cartoon or ink-wash animation in real-time. When users upload a video or record what they are seeing with the phone, the video-to-animation translator renders the live streaming video with cartoon or ink-wash animation style while maintaining the original contents. We formulate this task as video-to-video translation problem in the absence of any paired training examples, since the manual labeling of such paired video-animation data is cost-expensive and even unrealistic in practice. Technically, an unified unpaired video-to-video translator is utilized to explore both appearance structure and temporal continuity in video synthesis. As such, not only the visual appearance in each frame but also motion between consecutive frames are ensured to be realistic and consistent for video translation. Based on these technologies, our demonstration can be conducted on any videos in the wild and supports live video-to-animation translation, which engages users with the animated artistic expression of their life.

Live streaming has become a new form of entertainment, which attracts hundreds of millions of users worldwide. The huge amount of multimedia data in live streaming platforms creates tremendous opportunities for online advertising. However, existing state-of-the-art video advertising strategies (e.g., pre-roll and contextual mid-roll advertising) that rely on analyzing the whole video, are not applicable to live streaming videos. This paper describes a novel monetization framework, named LiveSense, for live streaming videos, which is able to display a contextually relevant ad at a suitable timestamp in a non-intrusive way. Specifically, given a live streaming video, we first employ a deep neural network to determine whether the current moment is appropriate for displaying an ad using the historical streaming data. Then, we detect a set of candidate ad insertion areas by incorporating image saliency, background map, and location priorities, so that the ad is displayed over the non-important area. We introduce three types of relevance metrics including textual relevance, global visual relevance and local visual relevance to select the contextually relevant ad. To minimize user intrusiveness, we initially display the ad at a non-important area. If the user is interested in the ad, we will show the ad in an overlaid window with a translucent background. Empirical evaluation on a real-world dataset demonstrates that our proposed framework is able to effectively display ads in live streaming videos while maintaining users' online experience.

Live streaming applications are becoming increasingly popular recently, and it exposes new technical challenges compared to regular video streaming. High video quality and low latency are two main requirements in live streaming scenarios. A live streaming application needs to make bitrate and target buffer level decisions as well as sets a continuous latency limit value to skip video frames. We formulate the live streaming task as a reinforcement learning problem with discrete-continuous hybrid action spaces, then propose a novel deep reinforcement learning (DRL) algorithm HD3 which can take hybrid actions to solve it. We compare HD3 with several state-of-the-art DRL algorithms on various network environments, and the simulation results show that HD3 can outperform all the other comparison schemes. We emphasize that HD3 generates a single agent which can perform well on different network conditions and video scenes.

Video compression algorithms result in a reduction of image quality, because of their lossy approach to reduce the required bandwidth. This affects commercial streaming services such as Netflix, or Amazon Prime Video, but affects also video conferencing and video surveillance systems. In all these cases it is possible to improve the video quality, both for human view and for automatic video analysis, without changing the compression pipeline, through a post-processing that eliminates the visual artifacts created by the compression algorithms. Generative Adversarial Networks have obtained extremely high quality results in image enhancement tasks; however, to obtain such results large generators are usually employed, resulting in high computational costs and processing time. In this work we present an architecture that can be used to reduce the computational cost and that has been implemented on mobile devices. A possible application is to improve video conferencing, or live streaming. In these cases there is no original uncompressed video stream available. Therefore, we report results using no-reference video quality metric showing high naturalness and quality even for efficient networks.

The live streaming is more challenging than on-demand streaming, because the low latency is also a strong requirement in addition to the trade-off between video quality and jitters in playback. To balance several inherently conflicting performance metrics and improve the overall quality of experience (QoE), many adaptation schemes have been proposed. Bitrate adaptation is one of the major solutions for video streaming under time-varying network conditions, which works even better combining with some latency control methods, such as adaptive playback rate control and frame dropping. However, it still remains a challenging problem to design an algorithm to combine these adaptation schemes together. To tackle this problem, we propose a hybrid control scheme for adaptive live streaming, namely HYSA, based on heuristic playback rate control, latency-constrained bitrate control and QoE-oriented adaptive frame dropping. The proposed scheme utilizes Kaufman's Adaptive Moving Average (KAMA) to predict segment bitrates for better rate decisions. Extensive simulations demonstrate that HYSA outperforms most of the existing adaptation schemes on overall QoE., 5 pages, ACMMM 2019

Digital gifting in live streaming, in which viewers buy digital gifts to reward the streamers, was worth over $200 million in 2018 in China and its growth has been accelerating. This paper explores what motivates people to tip and how it impacts interactions between viewers and streamers. Through a survey, we identified the main categories of viewers' tipping motivations. We found that viewers were motivated by the reciprocal acts of streamers, who would engage in various types of social interactions with tippers during the live streams. The styles of interactions and contents of live stream based on the tipping are differently influenced by the motivations of viewers and streamers. For example, viewers often tip large to attract attentions from the crowd or promote preferred live-streaming content. These findings provide more knowledge on the social interaction in live streaming platforms.