A Survey of Current YouTube Video Characteristics

Che, Xianhui; Ip, Barry; Ling, Lin

doi:10.1109/mmul.2015.34

Cited by 64 publications

(23 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various studies have explored the length, upload and access patterns, lifespan, ratings and comments, resolution and file size of YouTube videos to assess the extent to which they differ from traditional streaming content and help service providers optimize their networks (Cha et al, 2009; Che et al, 2015; Cheng et al, 2008). One study introduces a random prefix sampling approach to address the issue of unknown number of videos hosted on YouTube and estimates roughly 500 million videos by May 2011 (Zhou et al, 2011).…”

Section: Related Workmentioning

confidence: 99%

YouTube channels, uploads and views

Bärtl

2018

Convergence: The International Journal of Research into New Med

128

View full text Add to dashboard Cite

To this date, it is difficult to find high-level statistics on YouTube that paint a fair picture of the platform in its entirety. This study attempts to provide an overall characterization of YouTube, based on a random sample of channel and video data, by showing how video provision and consumption evolved over the course of the past 10 years. It demonstrates stark contrasts between video genres in terms of channels, uploads and views, and that a vast majority of on average 85% of all views goes to a small minority of 3% of all channels. The analytical results give evidence that older channels have a significantly higher probability to garner a large viewership, but also show that there has always been a small chance for young channels to become successful quickly, depending on whether they choose their genre wisely.

show abstract

Section: Related Workmentioning

confidence: 99%

YouTube channels, uploads and views

Bärtl

2018

Convergence: The International Journal of Research into New Med

128

View full text Add to dashboard Cite

show abstract

“…We have found from experience that this duration of video is sufficient to characterise the resource usage of the video. Additionally, a significant proportion of YouTube videos are less than 200 seconds in length [11]. …”

Section: A Methodologymentioning

confidence: 99%

Go Green with EnVI: the Energy-Video Index

Ejembi

Bhatti

2015

2015 IEEE International Symposium on Multimedia (ISM)

View full text Add to dashboard Cite

Video is the most prevalent traffic type on the Internet today. Significant research has been done on measuring user's Quality of Experience (QoE) through different metrics. We take the position that energy use must be incorporated into quality metrics for digital video. We present our novel, energyaware QoE metric for video, the Energy-Video Index (EnVI). We present our EnVI measurements from the playback of a diverse set of online videos. We observe that 4K-UHD (2160p) video can use ∼30% more energy on a client device compared to HD (720p), and up to ∼600% more network bandwidth than FHD (1080p), without significant improvement in objective QoE measurements.

show abstract

“…A reference channel with 3 sensors are located at Fpz. For the immersive VR stimuli, we chose a roller-coaster video from YouTube 360 [17] and displayed using Google's Cardboard VR technology [18]. The reason we chose a roller coaster video is to elicit a strong excitement reaction to the VR stimuli, in which the video has two specific exciting sections, one which is a drop from a high peak and another comprising a series of high speed 360-degrees turns.…”

Section: A Affective Virtual Realitymentioning

confidence: 99%

Deep Learning for Emotion Recognition in Affective Virtual Reality and Music Applications

2019

ijrte

View full text Add to dashboard Cite

This paper presents a deep learning approach to emotion recognition as applied to virtual reality and music predictive analytics. Firstly, it investigates the deep parameter tuning of the multi-hidden layer neural networks, which are also commonly referred to simply as deep networks that are used to conduct emotion detection in virtual reality (VR)- electroencephalography (EEG) predictive analytics. Deep networks have been studied extensively over the last decade and have shown to be among the most accurate methods for predictive analytics in image recognition and speech processing domains. However, most predictive analytics deep network studies focus on the shallow parameter tuning when attempting to boost prediction accuracies, which includes deep network tuning parameters such as number of hidden layers, number of hidden nodes per hidden layer and the types of activation functions used in the hidden nodes. Much less effort has been put into investigating the tuning of deep parameters such as input dropout ratios, L1 (lasso) regularization and L2 (ridge regularization) parameters of the deep networks. As such, the goal of this study is to perform a parameter tuning investigation on these deep parameters of the deep networks for predicting emotions in a virtual reality environment using electroencephalography (EEG) signal obtained when the user is exposed to immersive content. The results show that deep tuning of deep networks in VR-EEG can improve the accuracies of predicting emotions. The best emotion prediction accuracy was improved to over 96% after deep tuning was conducted on the deep network parameters of input dropout ratio, L1 and L2 regularization parameters. Secondly, it investigates a similar possible approach when applied to 4-quadrant music emotion recognition. Recent studies have been characterizing music based on music genres and various classification techniques have been used to achieve the best accuracy rate. Several researches on deep learning have shown outstanding results in relation to dimensional music emotion recognition. Yet, there is no concrete and concise description to express music. In regards to this research gap, a research using more detailed metadata on twodimensional emotion annotations based on the Russell’s model is conducted. Rather than applying music genres or lyrics into machine learning algorithm to MER, higher representation of music information, acoustic features are used. In conjunction with the four classes classification problem, an available dataset named AMG1608 is feed into a training model built from deep neural network. The dataset is first preprocessed to get full access of variables before any machine learning is done. The classification rate is then collected by running the scripts in R environment. The preliminary result showed a classification rate of 46.0%.

show abstract

A Survey of Current YouTube Video Characteristics

Cited by 64 publications

References 6 publications

YouTube channels, uploads and views

YouTube channels, uploads and views

Go Green with EnVI: the Energy-Video Index

Deep Learning for Emotion Recognition in Affective Virtual Reality and Music Applications

Contact Info

Product

Resources

About