Internet Video

Traffic monitoring reduces congestion, improves safety, and supports environmental sustainability. Real-time flow tracking, anomaly detection, and efficient management are key. Convolutional Neural Networks (CNNs) have become integral due to their compact size and easy deployment. However, their effectiveness depends heavily on the quality of the input data, especially image resolution. With highresolution cameras, especially 4K, balancing image quality, detection accuracy, and system efficiency is critical. We propose the Multi-Resolution Traffic Monitoring Dataset (MRTMD), which captures transport scenes at resolutions ranging from 2160p to 360p. This dataset serves as a benchmark for standard object detection models, enabling the development of more efficient and cost-effective traffic monitoring solutions. MRTMD will be freely available on GitHub, offering a valuable resource for researchers and practitioners. We evaluate leading object detection models-YOLOv9, YOLOv8, YOLOv7, Faster R-CNN, FCOS, SSD, and RT-DETR-across varied resolutions. Our analysis focuses on mean Average Precision (mAP), recall, and processing time. We also assess the accuracy of Number Plate Recognition (NPR) for tasks that require fine-grained detail extraction. Our findings show that detection performance typically varies within ±0.01 to ±0.03 in mAP and recall across resolutions, suggesting higher resolutions are not always advantageous. However, they remain crucial for tasks like NPR. The multi-resolution dataset enables a comprehensive evaluation of the trade-off between image quality and task performance. Ultimately, our analysis highlights the importance of resolution selection in large-scale deployments, informing system designers and policymakers. This dataset is a vital tool for balancing performance, cost, and practical constraints in real-world traffic monitoring. INDEX TERMS Dataset vehicle detection, high resolution image dataset, number plate recognition, computer vision.

HTTP Adaptive Streaming of video content is becoming an integral part of the Internet and accounts for the majority of today's traffic. Although Internet bandwidth is constantly increasing, video compression technology plays an important role and the major challenge is to select and set up multiple video codecs, each with hundreds of transcoding parameters. Additionally, the transcoding speed depends directly on the selected transcoding parameters and the infrastructure used. Predicting transcoding time for multiple transcoding parameters with different codecs and processing units is a challenging task, as it depends on many factors. This paper provides a novel and considerably fast method for transcoding time prediction using video content classification and neural network prediction. Our artificial neural network (ANN) model predicts the transcoding times of video segments for state of the art video codecs based on transcoding parameters and content complexity. We evaluated our method for two video codecs/implementations (AVC/x264 and HEVC/x265) as part of large-scale HTTP Adaptive Streaming services. The ANN model of our method is able to predict the transcoding time by minimizing the mean absolute error (MAE) to 1.37 and 2.67 for x264 and x265 codecs, respectively. For x264, this is an improvement of 22% compared to the state of the art.

Video streaming became an undivided part of the Internet. To efficiently utilise the limited network bandwidth it is essential to encode the video content. However, encoding is a computationally intensive task, involving high-performance resources provided by private infrastructures or public clouds. Public clouds, such as Amazon EC2, provide a large portfolio of services and instances optimized for specific purposes and budgets. The majority of Amazon's instances use x86 processors, such as Intel Xeon or AMD EPYC. However, following the recent trends in computer architecture, Amazon introduced Armbased instances that promise up to 40% better cost performance ratio than comparable x86 instances for specific workloads. We evaluate in this paper the video encoding performance of x86 and Arm instances of four instance families using the latest FFmpeg version and two video codecs. We examine the impact of the encoding parameters, such as different presets and bitrates, on the time and cost for encoding. Our experiments reveal that Arm instances show high time and cost saving potential of up to 33.63% for specific bitrates and presets, especially for the x264 codec. However, the x86 instances are more general and achieve low encoding times, regardless of the codec.

Named Data Networking (NDN), a data-centric enabled-cache architecture, as one of the candidates for the future Internet, has the potential to overcome many of the current Internet difficulties (\emph{e.g.}, security, mobility, multicasting). Influenced by using cache in intermediate equipment, NDN has gained attention as a prominent method of Internet content sharing. Managing the NDN caches and reducing the cache redundancy are the important goals in this paper. Our main contribution in this research is toward caching optimization in comparison with betweenness probabilistic in-network caching strategy. Therefore, with respect to combined impacts of long-term centrality-based metric and Linear Weighted Moving Average (LWMA) of short-term parameters such as user incoming pending requests and unique outgoing hit requests on caching management, a flexible probability caching strategy is proposed. Moreover, a simple Randomized-SVD approach is applied to combine averaged short-term and...

The goal of video summarization is to generate a shorter video sequence of a lengthy original sequence using only the key frames of the original sequence. We consider a video summarization scheme that generates a video summary that can be transmitted over an unreliable network such as the Internet with minimum distortion of the original video. We consider two methods of distortion measurement in our optimization scheme, and we apply the methods to a scenario in which feedback is available with the possibility of retransmitting lost packets. Simulation results showing the effectiveness of using the proposed schemes are presented.

This paper defines a Quality of Experience (QoE) parameter for multi-view video streaming service over Wireless Broadband Internet, and proposes cross-layer optimization (CLO) algorithm to maximize the QoE. The proposed CLO algorithm interacts with all layers from PHY layer to video layer. Under the time-varying wireless channel condition, it guarantees service continuity to the user by selecting frames to be transmitted. The CLO technique takes interview and the temporal significance of the multi-view video frames into consideration when deciding the Transmission Frame Set (TFS) and its Levels of FEC (LOF). Through simulation of the proposed technique, it is shown that the optimal quality of the multiview video can be achieved even during handover , using the minimum amount of resources. Visual quality of multi-view video streams is improved by about 15.3∼30.2%, especially up to 39.1∼65% during the handover. Moreover, continuity of the most important view has been increased by 20.1∼39.1% in comparison with conventional method.

Observation of surgical procedures performed by experts is extremely important for acquisition and improvement of surgical skills. Smart glasses are small computers, which comprise a head-mounted monitor and video camera, and can be connected to the internet. They can be used for remote observation of surgeries by video streaming. Although Google Glass is the most commonly used smart glasses for medical purposes, it is still unavailable commercially and has some limitations. This article reports the use of a different type of smart glasses, InfoLinker, for surgical video streaming. InfoLinker has been commercially available in Japan for industrial purposes for more than 2 years. It is connected to a video server via wireless internet directly, and streaming video can be seen anywhere an internet connection is available. We have attempted live video streaming of knee arthroplasty operations that were viewed at several different locations, including foreign countries, on a common web ...

By 2022, we expect video traffic to reach 82% of the total internet traffic. Undoubtedly, the abundance of video-driven applications will likely lead internet video traffic percentage to a further increase in the near future, enabled by associate advances in video devices' capabilities. In response to this ever-growing demand, the Alliance for Open Media (AOM) and the Joint Video Experts Team (JVET) have demonstrated strong and renewed interest in developing new video codecs. In the fast-changing video codecs' landscape, there is thus, a genuine need to develop adaptive methods that can be universally applied to different codecs. In this study, we formulate video encoding as a multi-objective optimization process where video quality (as a function of VMAF and PSNR), bitrate demands, and encoding rate (in encoded frames per second) are jointly optimized, going beyond the standard video encoding approaches that focus on rate control targeting specific bandwidths. More specifically, we create a dense video encoding space (offline) and then employ regression to generate forward prediction models for each one of the afore-described optimization objectives, using only Pareto-optimal points. We demonstrate our adaptive video encoding approach that leverages the generated forward prediction models that qualify for real-time adaptation using different codecs (e.g., SVT-AV1 and x265) for a variety of video datasets and resolutions. To motivate our approach and establish the promise for future fast VVC encoders, we also perform a comparative performance evaluation using both subjective and objective metrics and report on bitrate savings among all possible pairs between VVC, SVT-AV1, x265, and VP9 codecs. INDEX TERMS Video codecs, video signal processing, video coding, video compression, video quality, video streaming, adaptive video streaming, versatile video coding, AV1, HEVC.

3D video is a new form of video featuring omniview and immersive properties that are beneficial to many applications like telediagnosis, remote physiotherapy and elearning. In these applications, the data are always multi-modal and highly synchronized. For 2D video streaming, the MPEG group standardized Dynamic Adaptive Streaming over HTTP (DASH) with major transport formats, MPEG-2 TS and ISO Base Media File, including video, audio and subtitle formats. This paper presents OmniViewer, a multi-modal 3D streaming system based DASH, enhancing the versatility of the existing DASH standard and delivery infrastructure such as HTTP servers. Besides 3D video, OmniViewer supports multi-modal media in the DASH content for real-time media-synchronized interactive services. OmniViewer provides a complete multi-modal 3D video streaming solution from recording to rendering through DASH, which fills the gap between monoscopic 3D systems and 3D streaming systems.

Embedding new visible objects such as video or images intoMPEG video has many applications in newscasting, payper-view, Interactive TV and other distributed video applications. Because the embedded foreground content interfereswith the original motion compensation process ofthe background stream, we need to decode macroblocks inI and P frames via motion compensation and re-encodeall macroblocks with broken reference links via motion reestimation. Although previous work has explored DCTcompresseddomain ...

Over the years over-the-top (OTT) platforms such as Netflix and Hulu have attracted a lot of customers and caused serious disruptions in the media and advertising industry. This paper examines the underlying cause of such a disruption by using big data from one of the leading multi-system operators (MSO) in the US, which provides cable TV, phone and Internet services to multiple communities. More specifically, the dataset allows us to model the behaviour of 267,276 unique households, who collectively watched 270,718 unique programs and examine the effects of online video consumption, different genres, and use of additional cable services on cord shaving. The findings suggest the most important predictors of cord shaving are video on demand and time spent watching different genres, especially live sports and news. Finally, we discuss the practical implications of the research for content distributors, content providers, and advertisers.

This talk will discuss the complexities of designing news recommender system (RS) outputs such as newsletters and websites. There will be two main parts to the talk: (1) designing news RS outputs to balance between short-and longer-term objectives; and (2) creating bundled recommendations with the table d'hote ("host's table") approach. First, RS often optimize shortterm metrics that can be easily measured, but the strategic business goals of the system are typically realized over a longer period of time and are more difficult to measure, e.g., customer lifetime value (CLV), which is the discount sum of expected future cash flows due to the relationship. I will give empirical examples showing how optimizing some short-term news goals can harm the long-term goal. This hazard can be avoided by proposing and testing a causal model linking the shortand long-term goals through mediators. In the case of news RS, the key mediator is engagement, which is how news outputs create value for readers. I will establish short-term engagement metrics. The second part describes the table d'hote approach to create engaging RS outputs such as newsletters, which is motivated by the task of creating other "bundles" such as finedining experiences, musical concert set lists and art museum special exhibits. In all of these situations, the sequencing of items is an important consideration beyond more traditional RS goals such as coverage, novelty, diversity and serendipity. Table d'hote builds on communication theories and stratified sampling to create automated bundled news recommendations. It includes a typology of how news creates value for readers through different surveillance and serendipity experiences. I will discuss challenges for future research.

There has been a tremendous increase in internet users and enough bandwidth in recent years. Because Internet connectivity is so inexpensive, information sharing (text, audio, and video) has become more popular and faster. This video content must be examined in order to classify it for different purposes for users. Several machine learning approaches for video classification have been developed to save users time and energy. The use of deep neural networks to recognize human behavior has become a popular issue in recent years. Although significant progress has been made in the field of video recognition, there are still numerous challenges in the realm of video to be overcome. Convolutional neural networks (CNNs) are well-known for requiring a fixedsize image input, which limits the network topology and reduces identification accuracy. Despite the fact that this problem has been solved in the world of photos, it has yet to be solved in the area of video. We present a ten stacked three-dimensional (3D) convolutional network based on the spatial pyramidbased pooling to handle the input problem of fixed size video frames in video recognition. The network structure is made up of three sections, as the name suggests: a ten-layer stacked 3DCNN, DenseNet, and SPPNet. A KTH dataset was used to test our algorithms. The experimental findings showed that our model outperformed existing models in the area of video-based behavior identification by 2% margin accuracy.

Home users are increasingly acquiring, at lower prices, electronic devices such as video cameras, portable audio players, smartphones, and video game devices, which are all interconnected through the Internet. This increase in digital equipment ownership induces a massive production and sharing of multimedia content between these users. The supervised learning machine method Support Vector Machine (SVM) is vastly used in classification. It is capable of recognizing patterns of samples of predefined classes and supports multi-class classification. The purpose of this article is to explore the classification of multimedia P2P traffic using SVMs. To obtain relevant results, it is necessary to properly adjust the so-called Self C parameter. Our results show that SVM with linear kernel leads to the best classification results of P2P video with an F-Measure of 99% for C parameter ranging from 10 to 70 and to the best classification results of P2P file-sharing with an F-Measure of 98% for C parameter ranging from 30 to 70. We also compare these results with the ones obtained with Kolmogorov-Smirnov (KS) tests and Chi-square tests. It is shown that SVM with linear kernel leads to a better classification performance than KS and chi-square tests, which reached an F-Measure of 67% and 70% for P2P filesharing and P2P video, respectively, for KS test, and reached an F-Measure of 85% for both P2P file-sharing and P2P video for chi-square test. Therefore, SVM with linear kernel and suitable values for the Self C parameter can be a good choice for identifying encrypted multimedia P2P traffic on the Internet.

This article presents near misses as corrective and preventive measures to safety events. The article focuses on the risk factors of commercial motorcycling near misses, which we address by proposing a near miss detection framework based on a hybrid of YOLOv4-DeepSort and VGG16-BiLSTM models. We employed YOLOv4-DeepSort model for the detection and tracking tasks, and the tracked images and identity information were stored. The sequence of image was fetched into the VGG16-BiLSTM model for extraction of image feature information and near misses recognition respectively. Video streams of near miss datasets containing motorcycling in different scenes were collected for the experiment. We evaluate the proposed methods by testing 444 sequential video frames of motorcycling near misses in urban environment. The detection models achieved 96% accuracy for motorcycle, 89% for car, and 81% for person with lower false-positive rates on the test datasets while the tracking models achieved 34.3 MOTA on the test set and MOTP of 0.77. The results of the study indicate practicality for automatic detection of motorcycling near misses in urban environment, and it could assist in providing resourceful technical reference for analyzing the risk factors of motorcycling near misses. The research contributions are: (1) A hybrid of YOLOv4 and DeepSort model to enhance object detection and tracking in a complex environment and (2) A hybrid of YOLOv4 and DeepSort model to optimize the extraction of image feature information and near misses recognition respectively for overall system performance.

Super resolution is a technique that is getting increased attention. It was first deployed in high-end TVs to upconvert content, and now content providers are looking at how it can be used on the production side to guarantee the same high quality on all devices. Different techniques can be used to apply super resolution to SD and HD content for transmitting UHD resolution. During a recent trial with France TV at the French Open, SD content with a 16:9 aspect ratio was processed offline to create UHD content, using a hybrid pipeline of deep restoration and super resolution methods. A pretrained deep neural network was used to produce an intermediate upscaled and deblurred HD version, which was then passed on to a faster, less memory-intensive, machine-learning-based super resolution filter to produce the final UHD 4K resolution output. The result is a significantly better subjective viewing experience in terms of details and sharpness compared with classical filters used in live systems.

Super resolution is a technique that is getting increased attention. It was first deployed in high-end TVs to upconvert content, and now content providers are looking at how it can be used on the production side to guarantee the same high quality on all devices. Different techniques can be used to apply super resolution to SD and HD content for transmitting UHD resolution. During a recent trial with France TV at the French Open, SD content with a 16:9 aspect ratio was processed offline to create UHD content, using a hybrid pipeline of deep restoration and super resolution methods. A pretrained deep neural network was used to produce an intermediate upscaled and deblurred HD version, which was then passed on to a faster, less memory-intensive, machine-learning-based super resolution filter to produce the final UHD 4K resolution output. The result is a significantly better subjective viewing experience in terms of details and sharpness compared with classical filters used in live systems.

In this paper, we present a novel subjective quality model for online adaptive movie streaming service. The proposed model considers the Quality of Experience (QoE) of streaming video viewing as a cumulative evaluation process of consecutive segments that compose a story line. Under bandwidth constraint, streaming client may select lower-quality segment, pause playback for re-buffering, or both. The momentary QoE loss at these events are weighted by the content suspension level. Meanwhile such experience penalty remains to influence user's opinion for the rest of the service. If the picture becomes too noisy or the interruptions occur consistently, the user may stop watching. The proposed scheme includes two parts. First, a parametric model estimates the quality loss of a interfered segment based on its networklevel packet characteristics. Second, a cumulative function integrates the impact of streaming events. Both steps demand minimum computation and can be updated in real time. A subjective test to train and validate the proposed parametric model is designed and performed. This model is fundamentally different from all existing QoE assessment schemes in that temporally cumulative viewing experience of the users, instead of simple global statistics, is evaluated.

This research work designs and evaluates synchronized second screen interactive digital media using Moving Picture Experts Group Dynamic Adaptive Streaming over HTTP (MPEG-DASH). Second screen technology involves the use of a computing device, such as a smartphone or tablet, to provide an enhanced viewing experience for content on another device. The study design uses a structured format for defining MPEG-DASH inline events to trigger the second screen events. Furthermore, the ability for users to make choices that impact the story was implemented by generating several manifests and enabling the trigger manifest to refresh after each question. By performing a formal subjective user experiment, this study assesses the Quality of Experience (QoE) performance of the prototype. The results were positive and strongly indicate a high QoE for the entire system. A secondary result of our work is enhancing the QoE methodology for the assessment of second screen applications. The implemented system can easily be incorporated in the MPEG-DASH standard, allowing support for using the inline event for second screen triggering.

Information-Centric or Content-Centric Networking (ICN/CCN) is a promising novel network architecture that naturally integrates in-network caching, multicast, and multipath capabilities, without relying on centralized application-specific servers. Software platforms are vital for researching ICN/CCN; however, existing platforms lack a focus on extensibility and lightweight implementation. In this paper, we introduce a newly developed software platform enabling CCN, named Cefore. In brief, Cefore is lightweight, with the ability to run even on top of a resourceconstrained device, but is also easily extensible with arbitrary plugin libraries or external software implementations. For large-scale experiments, a network emulator (Cefore-Emu) and network simulator (Cefore-Sim) have also been developed for this platform. Both Cefore-Emu and Cefore-Sim support hybrid experimental environments that incorporate physical networks into the emulated/simulated networks. In this paper, we describe the design, specification, and usage of Cefore as well as Cefore-Emu and Cefore-Sim. We show performance evaluations of in-network caching and streaming on Cefore-Emu and content fetching on Cefore-Sim, verifying the salient features of the Cefore software platform.

Home users are increasingly acquiring, at lower prices, electronic devices such as video cameras, portable audio players, smartphones, and video game devices, which are all interconnected through the Internet. This increase in digital equipment ownership induces a massive production and sharing of multimedia content between these users. The supervised learning machine method Support Vector Machine (SVM) is vastly used in classification. It is capable of recognizing patterns of samples of predefined classes and supports multi-class classification. The purpose of this article is to explore the classification of multimedia P2P traffic using SVMs. To obtain relevant results, it is necessary to properly adjust the so-called Self C parameter. Our results show that SVM with linear kernel leads to the best classification results of P2P video with an F-Measure of 99% for C parameter ranging from 10 to 70 and to the best classification results of P2P file-sharing with an F-Measure of 98% for C parameter ranging from 30 to 70. We also compare these results with the ones obtained with Kolmogorov-Smirnov (KS) tests and Chi-square tests. It is shown that SVM with linear kernel leads to a better classification performance than KS and chi-square tests, which reached an F-Measure of 67% and 70% for P2P filesharing and P2P video, respectively, for KS test, and reached an F-Measure of 85% for both P2P file-sharing and P2P video for chi-square test. Therefore, SVM with linear kernel and suitable values for the Self C parameter can be a good choice for identifying encrypted multimedia P2P traffic on the Internet.

Adaptive bitrate selection adjusts the quality of HTTP streaming video to a changing context. A number of different schemes have been proposed that use buffer state in the selection of the appropriate video rate. However, models describing the relationship between video quality levels and buffer occupancy are mostly based on heuristics, which often results in unstable and/or suboptimal quality. In this paper, we present a QoE-aware video rate evolution model based on buffer state changes. The scheme is evaluated within a realworld Internet environment, where it is shown to improve the stability of the video rate. Up to 27% gain in average video rate can be achieved compared to the baseline ABR. The average throughput utilisation at a steady-state reaches 100% in some of the investigated scenarios.

As well known, when the input video is upscaled, the effectiveness of its transcoding and delivery may suffer. The encoded stream may not look sharp and use more bits than necessary. Then with adaptive streaming, extra streams may be added to reach such a maximum resolution and bitrate. The result is a significant waste of storage, bandwidth, and compute resources. In this paper, we explain the origins of this problem, survey existing methods for addressing it, and then propose our solution. Our proposed design incorporates a novel "true resolution" detection technique and a traditional CAE (context-aware encoding) encoding ladder generator. The CAE generator receives the detected "true resolution" of content as a limit for resolutions to include in the ladder. Such a limit enables all subsequent savings. We describe the details of our proposed resolution detection method, bring examples explaining how it works, and then study the performance of our proposed system in practice. Our study, performed using 500 video assets representing 120 hours of real-world production material, confirms the effectiveness of this technique. It shows that in many practical cases, the incoming content is, in fact, upscaled and that adding a "true resolution" detector to CAE brings very appreciable savings in bandwidth, storage, and compute costs.

We provide actionable strategies that teachers can follow when implementing asynchronous video communication and feedback in their own courses. These strategies are based on our own research over many years in the use of asynchronous video in online teaching, as well as our review of the literature, and are provided to foster greater discussion on the pedagogical strategies surrounding the use of video in online teaching.

The growth of the entertainment industry around the world may be seen in the creation of new genres and the influx of artists and musicians into this field. Every day, a large amount of music is generated and released. The classification of this music based on genres and the recommendation of music to users is a crucial task for various music streaming platforms. Many artificial intelligence methods have been created to overcome this. Inadequate data for training is one of the biggest issues when it comes to building machine learning algorithms. A certain dataset contains a large number of redundant features, which may lead the models to overfit. Data filtering could be used to solve this issue. On the GTZAN data for music genre classification, this article constructed numerous Artificial Intelligence (AI) models and used a data filtering strategy. This study does a comparative analysis and discusses the results. The models developed and evaluated are Naive Bayes,

Adaptive bitrate selection adjusts the quality of HTTP streaming video to a changing context. A number of different schemes have been proposed that use buffer state in the selection of the appropriate video rate. However, models describing the relationship between video quality levels and buffer occupancy are mostly based on heuristics, which often results in unstable and/or suboptimal quality. In this paper, we present a QoE-aware video rate evolution model based on buffer state changes. The scheme is evaluated within a realworld Internet environment, where it is shown to improve the stability of the video rate. Up to 27% gain in average video rate can be achieved compared to the baseline ABR. The average throughput utilisation at a steady-state reaches 100% in some of the investigated scenarios.

Analyzing and judging of captured and retrieved images of the targets from the surveillance video cameras for person re-identification have been a herculean task for computer vision that is worth further research. Hence, reidentification of single persons by locations based on single objects by locations (SOLO) model is proposed in this paper. To achieve the reidentification goal, we based the training of the re-identification model on synchronized stochastic gradient descent (SGD). SOLO is capable of exploiting the contextual cues and segmenting individual persons by their motions. The proposed approach consists of the following steps: (1) reformulating the person instance segmentation as: (a) prediction of category and (b) mask generation tasks for each person instance, (2) dividing the input person image into a uniform grids, i.e., G×G grid cells in such a way that a grid cell can predict the category of the semantic and masks of the person instances provided the center of the person falls into the grid cell and (3) conducting person segmentation. Discriminating features of individual persons are obtained by extraction using convolution neural networks. On person re-identification Market-1501 dataset, SOLO model achieved mAP of 84.1% and 93.8% rank-1 identification rate, higher than what is achieved by other comparative algorithms such as PL-Net, SegHAN, Siamese, GoogLeNet, and M 3 L (IBN-Net50). On person re-identification CUHK03 dataset, SOLO model achieved mAP of 82.1 % and 90.1% rank-1 identification rate, higher than what is achieved by other comparative algorithms such as PL-Net, SegHAN, Siamese, GoogLeNet, and M 3 L (IBN-Net50). These results show that SOLO model achieves best results for person re-identification, indicating high effectiveness of the model. The research contributions are: (1) Application of synchronized stochastic gradient descent (SGD) to SOLO training for person re-identification and (2) Single objects by locations using semantic category branch and instance mask branch instead of detect-then-segment method, thereby converting person instance segmentation into a solvable problem of single-shot classification.

Nowadays, Internet mobile becomes more popular and the number of mobile subscribers (MSs) is increasing very quickly. Internet MSs often access social networks such as Facebook/YouTube to share/download interesting pictures, voice, video and so on. For this reason, they occupy a lot of bandwidth in backbone transmission as well as server resources. With a traditional clients-server connection, servers easily suffered massive overload by a huge number of MSs that access service at the same time. In the century of green information technology, Content-Centric Networking (CCN) protocol has been proposed as the suitable solution for the above problem. The CCN uses named-data to push and cache the content to the edge gateways, so that highly popular content is generated once but can be consumed many times. In this paper, we present a solution for Long Term Evolution (LTE) mobile network using CCN protocol to cache contents in the E-UTRAN Node B (eNodeB). With OPNET Modeler simulation tool, we conduct realistic mobile networks with a huge number of mobility LTE MSs access to the same service on a single server. The simulation results show that all MSs requests are responded successfully while offloading server traffic significantly.

In recent years, image and video super-resolution have gained attention outside the computer vision community due to the outstanding results produced by applying deep-learning models to solve the super-resolution problem. These models have been used to improve the quality of videos and images. In the last decade, video-streaming applications have also become popular. Consequently, they have generated traffic with an increasing quantity of data in network infrastructures, which continues to grow, e.g., global video traffic is forecast to increase from 75% in 2017 to 82% in 2022. In this paper, we leverage the power of deep-learning-based super-resolution methods and implement a model for video super-resolution, which we call VSRGAN+. We train our model with a dataset proposed to teach systems for high-level visual comprehension tasks. We also test it on a large-scale JND-based coded video quality dataset containing 220 video clips with four different resolutions. Additionally, we pro...

Customer evaluations on social media may help us remain competitive and comprehend our business's target market. By analysing consumer evaluations, a business owner can identify common themes, pain points, and desired features or enhancements. By analysing customer feedback across multiple channels, such as social media, online reviews, and customer service interactions, businesses can rapidly identify any negative sentiment or potential brand damage. The contribution of our study is to evaluate the performance of the Naive Bayes method for classifying customer feedback on courier delivery services obtained via Twitter. The Naive Bayes algorithm is selected due to its simplicity, which facilitates efficient computation, suitability for large datasets, outstanding performance on text classification, and ability to manage high-dimensional data. In this investigation, the Naive Bayes classifier accuracy is 0.506, which is considered to be low. According to our findings, the irrelevant feature classification resulting in an error throughout the categorization process. A large number of data appearance characteristics that do not correspond to the testing data category have been identified as a result of this occurrence.

This article presents a novel, end-to-end, qualityaware optimization framework for multimedia clouds, where path selection mechanisms are exploited in conjunction with media optimization in order to support multimedia delivery in a quality-aware manner. As wireless data traffic worldwide is characterized by exponential growth, with the most prominent part being multimedia services, consumers get in the challenging position to compete for the limited wireless network resources. Cloud technologies and especially Software-Defined Networking is the perfect candidate technology in order to provide an elastic, dynamic provisioning of network resources that adapt to a highly changing environment, where application requirements and data volumes vary over time. This work combines the selection of the optimum path in the core network with quality-aware media adaptation based on the current conditions of the wireless access network. Thus the proposed framework achieves efficient network resources utilization in an end-to-end fashion.

This paper presents a quantitative analysis of a sample of six hundred audiovisual initiatives targeting the web as a distribution platform. The main purpose of this study involves using our sample to establish what we call unidirectional and interactive dimensions, based on the description of several indicators that will determine the nature of the services implemented and applications used in all the items analysed. These dimensions will be compared with two variables in order to analyse, firstly, the content lines, and secondly, whether or not the items represent the extension of conventional communication media (unidirectionality) on the web. The resulting data will show and specify the main differences between the dimensions of unidirectionality and interactivity with regard to Rebut /

Named Data Networking is a novel communication system, which relies upon caching of content to provide data on request. The currently used caching systems like LRU, which has been studied here, tend to under-utilize the cache-capacity. A new system of caching has been proposed here, it is probability based, two-step caching and as the name suggests, consists of incorporating probability in decision making of caching. This project uses an open-source implementation of NDN simulation called NdnSim and modifies its source-code to test its concepts.

This study shows how advertisers can leverage emotion and attention to engage consumers in watching Internet video advertisements. In a controlled experiment, the authors assessed joy and surprise through automated facial expression detection for a sample of advertisements. They assessed concentration of attention through eye tracking and viewer retention by recording zapping behavior. This allows tests of predictions about the interplay of these emotions and interperson attention differences at each point in time during exposure. Surprise and joy effectively concentrate attention and retain viewers. However, importantly, the level rather than the velocity of surprise affects attention concentration most, whereas the velocity rather than the level of joy affects viewer retention most. The effect of joy is asymmetric, with higher gains for increases than losses for decreases. Using these findings, the authors develop representative emotion trajectories to support ad design and testing.

There has been a tremendous increase in internet users and enough bandwidth in recent years. Because Internet connectivity is so inexpensive, information sharing (text, audio, and video) has become more popular and faster. This video content must be examined in order to classify it for different purposes for users. Several machine learning approaches for video classification have been developed to save users time and energy. The use of deep neural networks to recognize human behavior has become a popular issue in recent years. Although significant progress has been made in the field of video recognition, there are still numerous challenges in the realm of video to be overcome. Convolutional neural networks (CNNs) are well-known for requiring a fixedsize image input, which limits the network topology and reduces identification accuracy. Despite the fact that this problem has been solved in the world of photos, it has yet to be solved in the area of video. We present a ten stacked three-dimensional (3D) convolutional network based on the spatial pyramidbased pooling to handle the input problem of fixed size video frames in video recognition. The network structure is made up of three sections, as the name suggests: a ten-layer stacked 3DCNN, DenseNet, and SPPNet. A KTH dataset was used to test our algorithms. The experimental findings showed that our model outperformed existing models in the area of video-based behavior identification by 2% margin accuracy.

The popularity of the video services on the Internet has evolved various mechanisms that target the Quality of Experience (QoE) optimization of video traffic. The video quality has been enhanced through adapting the sending bitrates. However, rate adaptation alone is not sufficient for maintaining a good video QoE when congestion occurs. This paper presents a cross-layer architecture for video streaming that is QoE-aware. It combines adaptation capabilities of video applications and QoE-aware admission control to optimize the trade-off relationship between QoE and the number of admitted sessions. Simulation results showed the efficiency of the proposed architecture in terms of QoE and number of sessions compared to two other architectures (adaptive architecture and non-adaptive architecture).

With the inevitable dominance of video traffic on the Internet, providing perceptually good video quality is becoming a challenging task. This is partly due to the bursty nature of video traffic, changing network conditions and limitations of network transport protocols. This growth of video traffic has made Quality of Experience (QoE) of the end user the focus of the research community. In contrast, Internet service providers are concerned about maximizing revenue by accepting as many sessions as possible, as long as customers remain satisfied. However, there is still no entirely satisfactory admission algorithm for flows with variable rate. The trade-off between the number of sessions and perceived QoE can be optimized by exploiting the bursty nature of video traffic. This paper proposes a novel algorithm to determine the upper limit of the aggregate video rate that can exceed the available bandwidth without degrading the QoE of accepted video sessions. A parameter β that defines the exceedable limit is defined. The proposed algorithm results in accepting more sessions without compromising the QoE of ongoing video sessions. Thus it contributes to the optimization of the QoE-Session trade-off in support of the expected growth of video traffic on the Internet.

Transmission of video traffic over the Internet has grown exponentially in the past few years with no sign of waning. This increasing demand for video services has changed user expectation of quality. Various mechanisms have been proposed to optimise Quality of Experience (QoE) of end user's video. Studying these approaches are necessary for new methods to be proposed or combination of existing ones to be tailored. We discuss challenges facing the optimisation of QoE for video traffic in this paper. It surveys and classifies these mechanisms based on their functions. The limitation of each of them is identified and future directions are highlighted.

The emergence of video applications and video capable devices have contributed substantially to the increase of video traffic on Internet. New mechanisms recommending video rate adaptation towards delivering enhanced Quality of Experience (QoE) at the same time making room for more sessions. This paper introduces a cross-layer QoE-aware architecture for video traffic over the Internet. It proposes that video sources at the application layer adapt their rate to the network environment by controlling their transmitted bit rate dynamically; and the edge of network at the network layer protects the quality of the active video sessions by controlling the acceptance of new session through a video-aware admission control. In particular, it will seek the most efficient way of accepting new video session and adapting transmission rates to free up resources for more session while maintaining the QoE of active sessions. The proposed framework will contribute to the preparation for the extreme growth of video traffic in the foreseeable future. Simulation results show that the proposed cross-layer architecture guarantees the QoE for the admitted sessions and utilizes the link more efficiently comparing to the rate adaptation only architecture.

This paper defines a Quality of Experience (QoE) parameter for multi-view video streaming service over Wireless Broadband Internet, and proposes cross-layer optimization (CLO) algorithm to maximize the QoE. The proposed CLO algorithm interacts with all layers from PHY layer to video layer. Under the time-varying wireless channel condition, it guarantees service continuity to the user by selecting frames to be transmitted. The CLO technique takes interview and the temporal significance of the multi-view video frames into consideration when deciding the Transmission Frame Set (TFS) and its Levels of FEC (LOF). Through simulation of the proposed technique, it is shown that the optimal quality of the multiview video can be achieved even during handover , using the minimum amount of resources. Visual quality of multi-view video streams is improved by about 15.3∼30.2%, especially up to 39.1∼65% during the handover. Moreover, continuity of the most important view has been increased by 20.1∼39.1% in comparison with conventional method.

The current Internet model has proved more sustainable then the provisioned capacity at the time when the architecture was designed. The voluminous growth of traffic over the Internet has brought challenges for the exiting networking architecture. The information centric paradigm appears to offer efficient solution towards content dissemination model. It is a content-focused networking paradigm rather than host-to-host communication. Caching is one of the major components of information centric networks. This paper is intended to explore the impact of cache on critical attributes of networks. We have made a comparative analysis of in-network and edge network caching mechanism using network simulation. The results proved that in-network caching mechanism is far better than network edge caching with improved throughout, increase link capacity to avoid congestion.

This document presents a number of considerations regarding evaluating Information-Centric Networking (ICN) and sheds some light on the impact of ICN on network security. It also surveys the evaluation tools currently available to researchers in the ICN area and provides suggestions regarding methodology and metrics. Status of This Memo This document is not an Internet Standards Track specification; it is published for informational purposes. This document is a product of the Internet Research Task Force (IRTF). The IRTF publishes the results of Internet-related research and development activities. These results might not be suitable for deployment. This RFC represents the consensus of the <insert_name> Research Group of the Internet Research Task Force (IRTF). Documents approved for publication by the IRSG are not a candidate for any level of Internet Standard; see Section 2 of RFC 7841. Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at http://www.rfc-editor.org/info/rfc7945.

The popularity of the video services on the Internet has evolved various mechanisms that target the Quality of Experience (QoE) optimization of video traffic. The video quality has been enhanced through adapting the sending bitrates. However, rate adaptation alone is not sufficient for maintaining a good video QoE when congestion occurs. This paper presents a cross-layer architecture for video streaming that is QoE-aware. It combines adaptation capabilities of video applications and QoE-aware admission control to optimize the trade-off relationship between QoE and the number of admitted sessions. Simulation results showed the efficiency of the proposed architecture in terms of QoE and number of sessions compared to two other architectures (adaptive architecture and non-adaptive architecture).

There&#39;s a huge transition going on now. Whatever was over the air is increasingly moving onto fibre, and what was on the wires is now being delivered over the air. We are talking about data. A technology commentator Peter Cochrane, former CTO and head of research at British Telecom says &quot;Data downloads for the year 2013-2014 is about 17GB/ Month per user, and by 2017 that will go up to 70GB / Month per user. The sole reason behind this is exponential increase in video content over internet, as more people are working either at home or on the move&quot;. In broadband hotspots like Hong Kong, where they have 100Mbps services even in HOTEL ROOMS, people are no longer interested in watching TV or listening to their favorite radio over the air as everything is being put down fibre, that fascinates them more. Likewise, in UK there is a dedicated, reliable, fast broadband network. This all hails the need of more powerful techniques for video streaming, which give rise to a powerfu...

The Internet Engineering Task Force (IETF) has proposed the Guaranteed Service (GS) in the Integrated Services model with firm delay and bandwidth guarantees. In this paper, we study the effects of a token bucket traffic shaper at the source on the transmission characteristics of Motion Picture Experts Group (MPEG) compressed video streams over the GS. We develop an analytical model of the traffic shaper, and also carry out numerical simulation of the transmission performance using MPEG trace data from several different video sequences. The analytical model and the numerical simulation results are in excellent agreement. Our study provides a technique to determine the token bucket parameters that have to be specified while setting up a GS flow. Our study reveals that the traffic shaping process gives an end application a wide range of flexibility in controlling the delay. By choosing appropriate token bucket traffic shaper parameters, delay can be reduced significantly without the need to reserve costly network resources for less delay-sensitive applications, while time-critical applications may specify the required token bucket parameters for minimum delay.

Continuous improvements to the state of the art have made it easier to create, send and receive vast quantities of video over the Internet. Catalysed by these developments, video is now the largest, and fastest growing type of traffic on modern IP networks. In 2015, video was responsible for 70% of all traffic on the Internet, with an compound annual growth rate of 27%. On the other hand, concerns about the growing energy consumption of ICT in general, continue to rise. It is not surprising that there is a significant energy cost associated with these extensive video usage patterns. In this thesis, I examine the energy consumption of typical video configurations during decoding (playback) and encoding through empirical measurements on an experimental test-bed. I then make extrapolations to a global scale to show the opportunity for significant energy savings, achievable by simple modifications to these video configurations. Based on insights gained from these measurements, I propose...

Due to the significant increase in the last five years of audiovisual content distribution over the web, this paper is focused on a study aimed at the description and classification of a wide sample of audiovisual initiatives whose access is carried out by means of the World Wide Web. The purpose of this study is to promote the debate concerning the different names of these incipient