Large Number Of Videos Research Articles

A method to increase video streaming speed and efficiency of storage video using multithreading and Random Access Memory

There are numerous vital applications for online video storage and network transmission today. With exceptional scalability and browser and device compatibility, Hypertext Transfer Protocol (HTTP) live streaming (HLS) is today's most popular video and audio transmission protocol. In the article, the authors proposed a method of using multithreading and Random Access Memory (RAM) to store and transmit videos to help the server achieve higher speed, improve video transmission speed and video storage efficiency. The results of this research will help multimedia server developers evaluate the appropriate balance of resource usage while still ensuring the quality of online video transmission with a large number of videos and viewers.

The Quality and Reliability of Information in YouTube Videos on Occupational Health and Safety.

Education is very important to prevent occupational injuries and accidents, which are almost all completely preventable. The aim of this study was to evaluate training videos on this subject on the YouTube platform. Six search terms related to occupational health and safety (OHS) were scanned on May 31, 2021. After the application of exclusion criteria, a total of 176 videos were included for final analysis using the parameters of country origin, source of the video, content, number of views, comments, likes, dislikes, and video duration. The Global Quality Scale (GQS) and modified DISCERN tools were used to evaluate the quality and reliability of the videos in this analytical cross-sectional study. According to the GQS score, 111 (63.1%) videos were of low quality. Statistically significant differences were found between the low-, moderate-, and high-quality groups with respect to video length, likes, dislikes, comments, likes per day, dislikes per day, comments per day, video category, and the DISCERN scores (P < 0.05). The vast majority of videos contained low-quality information. A large number of videos were uploaded on OHS content from independent users and the USA. There is a clear need for professionals to play a more active role in uploading and sharing high-quality information on Internet platforms on the subject of OHS.

Youth Subculture Performance from the Perspective of Emotional Communication——A Research Based on the Phenomenon of “Meaningless words”

In the context of "Traffic is King ", a large number of videos and content with empty meaning, lack of information and striving for traffic are flooded with the network. In the social platform, the audience has created many language expressions rich in "meme", such as "meaningless words". As an important part of youth subculture, "meaningless words" mainly creates characters through collages, spoofs, parodies and other forms, and its expression is more direct and simple. It is an important way of satirizing and roast on the social media platform. Why do contemporary youth produce "meaningless words"? Why does "meaningless words" spread so widely? This paper will explore the production mechanism of "meaningless words" from the perspective of emotional communication.

MULTIMEDIA SUMMARIZATION USING SPACY

The main objective of this paper is to propose a summarizing system based on natural language processing to summarize the data that can be in either video, audio or textual format. Video summarization is to generate a short summary of the content of a video by selecting and presenting the most informative or interesting materials for potential users. With the arrival of high-speed Internet and low-cost storage, the amount of data has been generated at a rocket pace, most of it is in the form of visual or video data. This makes them an important information dispensing tool. Searching and analyzing such a large number of videos is an extremely tedious task Unlike images, where information may be gathered from a single frame, videos need the user to watch the full content, regardless of duration, which presents a challenge to information extraction. Similarly, we save time by extracting crucial information from audio recordings and text. The system focuses on the development of a module using Natural Language Processing with python to summarize the YouTube videos, audio files and textual data. KEYWORDS - Multimedia, Natural Language Processing, Spacy, Summarization, Tokenization

Timeline Binder: A Caption-based Approach for Creating Semantic Video Timeline

Video is an important type of multimedia which has myriad characteristics such as richer content and little prior structure. Major applications such as surveillance, medicine, education, entertainment, and sports heavily use videos. Internet protocol video is dominating the World’s Internet traffic. Large number of videos are made and uploaded daily. Looking for topics of one’s interest in these videos is a cumbersome task. Mostly, people are only interested in small portions of videos as opposed to the whole video. It is better to look for small but related segments of interest. Our proposed system Timeline Binder is an attempt to tackle this problem. For Timeline Binder, we have created our own dataset of online videos with subtitles. Based on those subtitles’ files, a concept vector is obtained. If a concept vector has a match for a user’s query tag, then timed segments for all related videos to that matched concept in the concept vector are combined into a single timeline thus making all related segments into a single video. In this research, we present performance measurements between Timeline Binder and other available applications such as Google2SRT and NR8. The results show that Timeline Binder performs better than Google2SRT and NR8.

Social Media Videos on Contraceptive Implants: An Assessment of Video Quality and Reliability

There are many videos on contraception on YouTube and TikTok, and women around the world use video platforms extensively to seek health information from videos of uncertain reliability and quality. The main objective of this research was to evaluate the scientific reliability and quality of the most popular social media videos about contraceptive subdermal implants. TikTok and YouTube videos were assessed separately by two gynecologists. The video quality was evaluated according to the DISCERN score, modified DISCERN score, and Global Quality Scale (GQS) score. The study included the 100 most popular videos, 44 of which were uploaded by health care professionals. The median DISCERN score for videos shared by health care professionals was 50 (range 15-75), whereas it was 22 (range 15-56) for videos shared by independent users (P < .001). The median modified DISCERN score was 4 (min-max 0-5), and the median GQS score was 4 (min-max 1-5) for the videos uploaded by health care professionals. The modified DISCERN and GQS scores of TikTok videos were lower than those of YouTube videos (both P < .001). Social media videos about contraceptive subdermal implants are popular and have been viewed tens of thousands of times, but the overall medical validity, which we evaluated according to the scoring systems, was poor. Despite the large number of videos, the content's quality and reliability are quite limited. There is a need to prioritize the education provided to patients by gynecologists, as well as to create high-quality content for YouTube and other similar platforms.

SP11.8 Deep learning for performance assessment of laparoscopic sleeve gastrectomy (LSG): validation against objective structured assessment of technical skills (OSATS) and clinical outcomes

Abstract Background Current methods of surgeon performance assessment are unstructured, labour-intensive and subjective. Machine learning could provide rapid, automated and subjective assessment. We design a custom deep learning model to output performance metrics from videos of LSG and validate these against OSATS and clinical outcomes. Methods A novel dataset of 3210 images from videos of LSG was annotated to train an instrument segmentation framework based on Mask R-CNN, a state-of-the-art deep learning algorithm. Spatial outputs of detected instruments were used to determine workflow through a Markov chain and secondly to determine instrument metrics for assessment of surgical performance. Metrics including time spent in stage and instrument time, distance, smoothness, concentration, and efficiency were then validated against OSATS ratings from 2 independent scorers and clinical outcomes from 35 LSGs. Results Grasper and ligasure efficiency and smoothness were higher in videos rated within the top quartile by OSATS score when compared to those in the bottom quartile and when comparing consultants to trainees. Videos in the top quartile for grasper smoothness had higher total OSATS scores (17.31±2.27) compared to those in the lowest quartile (15.69±2.93) and across all 5 individual OSATS domains in addition to higher percentage of total weight loss at 3, 6, and 12 months. Conclusions This study demonstrates how vision based deep learning can be used for surgeon performance assessment. Future work will aim to validate this against a large number of videos, expand upon the range of metrics produced, and look towards the steps needed for clinical translation.

Ideological and political education system based on emotion analysis in large‐scale online education

The amount of data on the Internet is showing an explosive growth trend. The recommendation system can help users find the resources they need from a large number of videos, which has become an urgent problem to improve the effectiveness of online education. With the rapid development of Internet of Things technology, the timeliness of information collection and processing has been further improved. This paper constructs a multi‐information fusion sequence recommendation system for ideological and political online education. Specifically, the facial video information is collected by the camera, and these videos are delivered to the server. We introduce gate recurrent unit (GRU) to analyze the human expression state. By exploiting human emotion during online ideological and political video learning, we design a novel sequence recommendation system in which human emotional states and related videos are fused based on the multi‐head attention mechanism. Experimental results show that video recommendation performance can be effectively improved by introducing emotional information.

Auto Deep Learning-based Automated Surveillance Technique to Recognize the Activities in the Cyber-Physical System

In recent days, the Internet of Things (IoT) plays a significant role and increasing in rapid usage in various applications. As IoT is being developed for cyber-physical systems in the specific domain of e-health care, military, etc. Based on real-time applications, security plays a vital role in certain activities in educational institutions. In the institutions, there are multiple videos are collected and stored in the data repositories. Those datasets are developed specifically for certain activities and no other datasets are developed for academic activities. As there is a large number of videos and images are collected and considered, advanced technologies like, deep learning and IoT are used to perform certain tasks. In this paper, a Auto Deep learning-based Automated Identification Framework (DLAIF) is proposed to consider and reconsider the activities based on image pre-processing, model can be trained through the proposed GMM model and then predication to make an effective surveillance process based on HMM. This proposed process makes to recognize the activities through EM and log Likelihood for cyber-physical systems. In the performance analysis, the proposed model efficiency can be determined through Accuracy detection, False Positive rate and F1 Score requirement. Then calculating the accuracy is more effective for the proposed model compared to other existing models such as BWMP and LATTE.

Video Transcript Summarizer

In today’s world, a large number of videos are uploaded in everyday, which contains information about something. The major challenge is to find the right video and understand the correct content, because there are lot of videos available some videos will contain useless content and even though the perfect content available that content should be required to us. If we not found right one it wastes your full effort and full time to extract the correct usefull information. We propose an innovation idea which uses NLP processing for text extraction and BERT Summarization for Text Summarization. This provides a video main content in text description and abstractive summary, enabling users to discriminate between relevant and irrelevant information according to their needs. Furthermore, our experiments show that the joint model can attain good results with informative, concise, and readable multi-line video description and summary in a human evaluation.

Advanced Evaluation Method for Video Retrieval System

n the past few years, the multimedia storage has increased and costs of storing multi-media data has become cheaper. Which is why, there are large numbersof videos that are available in the video repositories. With development of the multi-media data types and available bandwidths.The proposed method transferseach video of database into scenes using color histogrambased scene change detection algorithm and key framesare extracted. For key frames multiple features areobtained using straight forward rules. A new framework depending on CNNs (convolutional neural networks) is recommended to perform the content based video retrieval with the less storage cost also with higher search capability. The recommended framework subsists of extraction algorithm with respect to key-frame and the feature collection strategies. Particularly, the extraction algorithm of key-frame takes benefit of clustering idea; so in that case excessive information is taken out from video data and also the storage cost is highly shortened. This work present a method uses the extracted features with convolutional neural network (CNN) for classification tasks. In this research paper, different types of videos will be used and the important features will be extracted using SIFT, after that we use the deep learning(CNN) process will be performed. The experimental results showed the efficiency and effectiveness of suggested approach

Cooperative caching for scalable video coding using value-decomposed dimensional networks

Scalable video coding (SVC) has been widely used in video-on-demand (VOD) service, to efficiently satisfy users' different video quality requirements and dynamically adjust video stream to time-variant wireless channels. Under the 5G network structure, we consider a cooperative caching scheme inside each cluster with SVC to economically utilize the limited caching storage. A novel multi-agent deep reinforcement learning (MADRL) framework is proposed to jointly optimize the video access delay and users' satisfaction, where an aggregation node is introduced helping individual agents to achieve global observations and overall system rewards. Moreover, to cope with the large action space caused by the large number of videos and users, a dimension decomposition method is embedded into the neural network in each agent, which greatly reduce the computational complexity and memory cost of the reinforcement learning. Experimental results show that: 1) the proposed value-decomposed dimensional network (VDDN) algorithm achieves an obvious performance gain versus the traditional MADRL; 2) the proposed VDDN algorithm can handle an extremely large action space and quickly converge with a low computational complexity.

A Non-Invasive Method for Measuring Bubble Column Hydrodynamics Based on an Image Analysis Technique

Bubble size and its distribution are the important parameters which have a direct impact on mass transfer in bubble column reactors. For this, a new robust image processing technique was presented for investigating hydrodynamic aspects and bubble behavior in real chemical or biochemical processes. The experiments were performed in a small-scale bubble column. The study was conducted for the wide range of clear liquid heights and superficial gas velocities. However, a major challenge in image analysis techniques is identification of overlapping or cluster bubbles. This problem can be overcome with the help of the proposed algorithm. In this respect, large numbers of videos were recorded using a high-speed camera. Based on detailed experiments, the gas–liquid dispersion area was divided into different zones. A foam region width was found as inversely proportional to the clear liquid height. An entry region width was found as directly proportional to the clear liquid height. Hydrodynamic parameters, including gas holdup, bubble size distribution, and Sauter mean bubble diameter were evaluated and compared for different operating conditions. The gas holdup was calculated from both height measurement and pixel intensity methods, and it was found to be indirectly proportional to clear liquid height. Bubble sizes affect the bubble column performance; therefore, bubbles are tracked to calculate the bubble size distribution. Experimental results proved that the proposed scheme is robust.

Identifying Popular Frogs and Attractive Frog Calls from YouTube Data

Public interest in and preferences for certain species can sometimes provide an opportunity for conservation and management. Here, we attempted to identify ‘popular’ anurans from YouTube data. In addition, the attractiveness of anuran advertisement-calling sounds were analyzed using acoustic data. By searching YouTube with the search term ‘frog calling’, 250 videos were selected. Of these, 174 videos could be classified according to species; these videos aided in extracting clean calling sounds, free from the overlapping calls of other male frogs, as well as other sounds. To assess the interests and preferences of viewers for different species, the numbers of videos, view counts, ‘likes,’ and ‘dislikes’ were recorded. From the videos, the calls of 78 species belonging to 17 families were identified. Viewer interest was highest for the Hylidae and Ranidae species, which are often discoverable in the field. In addition, invasive frogs had large numbers of videos and large numbers of ‘likes.’ People tended to prefer frogs calling with lower dominant frequencies. However, there were few videos on endangered species, and these garnered relatively less interest than other species. To manage and conserve invasive or endangered frog species, there is a need to increase ecological understanding by adjusting species awareness and charisma.

Attribute based spatio-temporal person retrieval in video surveillance

Many venues, such as airports, railway stations, and shopping malls, have video surveillance systems for security and monitoring. However, searching for and retrieving people based on attribute descriptions in a large number of videos is difficult, particularly with weather variations and crowded places. Most of the existing attribute-based person retrieval systems consist of two main modules: object detection and person attribute recognition. The common drawbacks of object detection in the existing methods are false-positive, missing detection, and multi bounding boxes for the same object. Moreover, attribute recognition algorithms suffer from low accuracy for a single attribute classifier, while attributes error spread in the cascading multi-attribute classifier. This paper overcomes these issues by applying the ByteTrack algorithm instead of object detection to exploit the person's spatio-temporal information and generate a tube that maintains all the boxes that include the objects and associates high and low score boxes of the objects without raising false positive detection. Also, linking each person bounding boxes together results in more accurate attributes recognition than defining the attributes of each bounding box separately. Moreover, the proposed algorithm merges between selected predictions of two attribute recognition algorithms to improve the recognition performance. An extensive empirical evaluation was carried out on the SoftBioSearch database. The simulation results reveal that the proposed retrieval algorithm provides effective retrieval performance that exceeds the best conventional method by 14%.

A Violent Video Detection Method Based on Image Semantic Segmentation

With the rapid growth of the network user base and the number of short videos, a large number of videos related to terrorism and violence have emerged in the Internet, which has brought great challenges to the governance of the network environment. At present, most short-video platforms still adopt manual-review and user-report mechanisms to filter videos related to terrorism and violence, which cannot adapt to the development trend of short-video business in terms of recognition accuracy and timeliness. In the single-mode recognition method of violent video, this paper mainly studies the scene recognition mode. Firstly, the U-Net network is improved with the SE-block module. After pretraining on the Cityscapes dataset, semantic segmentation of video frames is carried out. On this basis, semantic features of scenes are extracted using the VGG16 network loaded with ImageNet pretraining weights. SE-U-Net-VGG16 scene recognition model is constructed. The experimental results show that the prediction accuracy of SE-U-Net model is much higher than that of the FCN model and U-Net model. SE-U-Net model has significant advantages in the modal research of scene recognition.

Improving the model of object detection on aerial photographs and video in unmanned aerial systems

This paper considers a model of object detection on aerial photographs and video using a neural network in unmanned aerial systems. The development of artificial intelligence and computer vision systems for unmanned systems (drones, robots) requires the improvement of models for detecting and recognizing objects in images and video streams. The results of video and aerial photography in unmanned aircraft systems are processed by the operator manually but there are objective difficulties associated with the operator’s processing of a large number of videos and aerial photographs, so it is advisable to automate this process. Analysis of neural network models has revealed that the YOLOv5x model (USA) is most suitable, as a basic model, for performing the task of object detection on aerial photographs and video. The Microsoft COCO suite (USA) is used to train this model. This set contains more than 200,000 images across 80 categories. To improve the YOLOv5x model, the neural network was trained with a set of VisDrone 2021 images (China) with the choice of such optimal training parameters as the optimization algorithm SGD; the initial learning rate (step) of 0.0005; the number of epochs of 25. As a result, a new model of object detection on aerial photographs and videos with the proposed name VisDroneYOLOv5x was obtained. The effectiveness of the improved model was studied using aerial photographs and videos from the VisDrone 2021 set. To assess the effectiveness of the model, the following indicators were chosen as the main indicators: accuracy, sensitivity, the estimation of average accuracy. Using a convolutional neural network has made it possible to automate the process of object detection on aerial photographs and video in unmanned aerial systems.

Analysis of Informative Content on Cerebral Palsy Presented in Brazilian-Portuguese YouTube Videos

Aims: To describe the characteristics of the most accessed YouTube videos in Brazilian-Portuguese on cerebral palsy (CP), and to analyze content of informational videos about this topic. Methods: This was a cross-sectional study. Searching on YouTube website was conducted by two independent examiners between November and December 2019, using the keywords “Paralisia Cerebral” sorted by videos’ number of views. Videos that did not present content related to CP or duplicate videos were excluded. The interaction parameters and content characteristics of the included videos were extracted. To access the trustworthiness and quality of informational videos, the modified Discern checklist and the Global Quality Score was used. Results: Following the eligibility criteria 90 videos were included. Fifty-three (53) were classified as experiential videos and 37 as informational videos. Informational videos presented multi-topics about different aspects of CP. This group of videos presented moderate trustworthiness due to the lack of scientific evidence content. Informational videos had good quality and generally good flow. Conclusion: YouTube presented a large number of videos about CP in Brazilian-Portuguese. Informational videos are useful for patients and healthcare providers; however, it is necessary to included information about scientific evidence, as a strategy to facilitate and promote knowledge translation.

The Relevance of Videos as a Practical Tool for Communication and Dissemination in Horizon2020 Thematic Networks

Thematic networks (TNs) are Horizon2020 projects promoted by the European Innovation Partnership Agricultural Productivity and Sustainability (EIP-AGRI) that collect existing knowledge and best practices on a given theme in agriculture and forestry. Their purpose is to make these available in easily understandable formats for users such as farmers/foresters and advisors. Videos are a powerful tool for communication and dissemination about project outcomes; thus, they are frequently used as part of project activities. Despite the large number of videos that have been produced by TNs, only a few have reached a high number of views. Hence, most have not reached a wide public, decreasing the impact of TNs’ outputs. This paper analyzes several characteristics of TN videos in order to suggest best practices. That is, we investigated whether each characteristic influences the number of views or average percentage watched. Using the analysis, we suggest several best practices. For example, future TNs should show practical content prioritizing farmers as a speaker and translate videos in more than one language. Better involvement of project partners increases the contact with farmers; thus, outcomes are more easily spread, and the chance of being applied increases. To increase the visibility of videos, a user-friendly digital knowledge platform linked to demonstration activities and peer-to-peer exchange will improve the sharing of knowledge. This will further enhance a TN’s impact and sustainability in agricultural and forestry innovation in the longer term.

Accelerating temporal action proposal generation via high performance computing

Temporal action recognition always depends on temporal action proposal generation to hypothesize actions and algorithms usually need to process very long video sequences and output the starting and ending times of each potential action in each video suffering from high computation cost. To address this, based on boundary sensitive network we propose a new temporal convolution network called Multipath Temporal ConvNet (MTN), which consists of two parts i.e. Multipath DenseNet and SE-ConvNet. In this work, one novel high performance ring parallel architecture based on Message Passing Interface (MPI) is further introduced into temporal action proposal generation, which is a reliable communication protocol, in order to respond to the requirements of large memory occupation and a large number of videos. Remarkably, the total data transmission is reduced by adding a connection between multiple computing load in the newly developed architecture. It is found that, compared to the traditional Parameter Server architecture, our parallel architecture has higher efficiency on temporal action detection task with multiple GPUs, which is suitable for dealing with the tasks of temporal action proposal generation, especially for large datasets of millions of videos. We conduct experiments on ActivityNet-1.3 and THUMOS14, where our method outperforms other state-of-art temporal action detection methods with high recall and high temporal precision. In addition, a time metric is further proposed here to evaluate the speed performance in the distributed training process.