• Journal of Broadcast Engineering
• /
• v.27 no.6
• /
• pp.897-905
• /
• 2022

Research for video key frame detection has been actively conducted in the fields of computer vision. Recently with the advances on deep learning techniques, performance of key frame detection has been improved, but the various type of video content and complicated background are still a problem for efficient learning. In this paper, we propose a novel method for key frame detection, witch utilizes contrastive learning and memory bank module. The proposed method trains the feature extracting network based on the difference between neighboring frames and frames from separate videos. Founded on the contrastive learning, the method saves and updates key frames in the memory bank, witch efficiently reduce redundancy from the video. Experimental results on video dataset show the effectiveness of the proposed method for key frame detection.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.4
• /
• pp.65-77
• /
• 2013

Recently the ratio of the Internet traffic generated by video streaming applications including video-on-demand (VOD) is getting higher and higher, while P2P-based naive content distribution has been the main source of the Internet traffic in the past. As a result, the paradigm of cooperatively distributed systems (e.g., P2P) is changing to support streaming applications. Most P2P assisted approaches for video streaming today are based on Bit Torrent thanks to its simplicity of implementation and easy adaptability. They, however, have immanent vulnerability to free-riding inherited from Bit Torrent, which inevitably hurts their performance under limited system resources with free-riding. This paper studies the weakness to free-riding of existing Bit Torrent-based video streaming applications and investigates the adaptability of T-Chain (which was originally designed to prevent free-riding in cooperatively distributed systems) to video streaming applications. Our experiment results show that the video streaming approach based on T-Chain outperforms most existing Bit Torrent-based ones by 60% on average under limited system resources with free-riding.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.5 no.2
• /
• pp.152-157
• /
• 2010

The development of computer and internet has introduced various types of media - such as, image, audio, video, and voice - to the traditional text-based information. However, most of the information retrieval systems are based only on text, which results in the absence of ability to use available information. By utilizing the available media, one can improve the performance of search system, which is commonly called content-based retrieval and content-based image retrieval system specifically tries to incorporate the analysis of images into search systems. In this paper, a content-based image retrieval system using HSI color space, ART2 algorithm, and SOM algorithm is introduced. First, images are analyzed in the HSI color space to generate several sets of features describing the images and an SOM algorithm is used to provide candidates of training features to a user. The features that are selected by a user are fed to the training part of a search system, which uses an ART2 algorithm. The proposed system can handle the case in which an image belongs to several groups and showed better performance than other systems.

• Journal of Broadcast Engineering
• /
• v.25 no.4
• /
• pp.518-527
• /
• 2020

360 VR technology is used not only in movies, but also in performing arts such as music, theater, dance, and so on due to the characteristics of immersion and presence. The technology allows the audience can be perceived a feel of participation in a story. This study is conducted an analysis of the techniques of 360 video shooting in order to find the answers of the following questionaries: how to make viewers enhance to a better understanding of a space, how to make the viewer feel comfortable ceding control of the experience, how to generate greater empathy with a 360 video. Thirty cases were analysed 360-degree videos of live performances performed on stage among 360-degree images of music performance content shared on Youtube from 2015 to 2020. The result shows that live performances are performed with the audience, so the stage shape and the layout of the audience seats are preferred to the characteristics of the performance. It was also shown that directing using a 360 VR camera was also greatly affected by the stage and audience placement. The stage is manly classified into three types, and the camera layout and characteristics mainly used are organized according to the number of 360 VR cameras, whether fixed or mobile cameras are used.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.4
• /
• pp.121-131
• /
• 2022

We look at the initial stage of Meta (previous Facebook)'s new metaverse platform and investigate its platform design in pre-launch and ignition life stage. From the Rocket Model (RM)'s theoretical logic, the results reveal that Meta firstly focuses on investing in key content developers by acquiring virtual reality (VR), video, music content firms and offering production support platform of the augmented reality (AR) content, 'Spark AR' last three years (2019~2021) for attracting high-potential developers and users. In terms of three matching criteria, Meta develops an Artificial Intelligence (AI) powered translation software, partners with Microsoft (MS) for cloud computing and AI, and develops an AI platform for realistic avatar, MyoSuite. In 'connect' function, Meta curates the game concept submitted by game developers, welcomes other game and SNS based metaverse apps, and expands Horizon Worlds (HW) on VR devices to PCs and mobile devices. In 'transact' function, Meta offers 'HW Creator Funding' program for metaverse, launches the first commercialized Meta Avatar Store on Meta's conventional SNS and Messaging apps by inviting all fashion creators to design and sell clothing in this store. Mata also launches an initial test of non-fungible token (NFT) display on Instagram and expands it to Facebook in the US. Lastly, regarding optimization, especially in the face of recent data privacy issues that have adversely affected corporate key performance indicators (KPIs), Meta assures not to collect any new data and to make its privacy policy easier to understand and update its terms of service more user friendly.

• ETRI Journal
• /
• v.36 no.4
• /
• pp.662-672
• /
• 2014

Human action recognition is used in areas such as surveillance, entertainment, and healthcare. This paper proposes a system to recognize both single and continuous human actions from monocular video sequences, based on 3D human modeling and cyclic hidden Markov models (CHMMs). First, for each frame in a monocular video sequence, the 3D coordinates of joints belonging to a human object, through actions of multiple cycles, are extracted using 3D human modeling techniques. The 3D coordinates are then converted into a set of geometrical relational features (GRFs) for dimensionality reduction and discrimination increase. For further dimensionality reduction, k-means clustering is applied to the GRFs to generate clustered feature vectors. These vectors are used to train CHMMs separately for different types of actions, based on the Baum-Welch re-estimation algorithm. For recognition of continuous actions that are concatenated from several distinct types of actions, a designed graphical model is used to systematically concatenate different separately trained CHMMs. The experimental results show the effective performance of our proposed system in both single and continuous action recognition problems.

• The Journal of Information Systems
• /
• v.32 no.3
• /
• pp.1-17
• /
• 2023

Purpose The proliferation of data on the internet has created a need for innovative methods to analyze user satisfaction data. Traditional survey methods are becoming inadequate in dealing with the increasing volume and diversity of data, and new methods using unstructured internet data are being explored. While numerous comment-based user satisfaction studies have been conducted, only a few have explored user satisfaction through video and audio data. Multimodal sentiment analysis, which integrates multiple modalities, has gained attention due to its high accuracy and broad applicability. Design/methodology/approach This study uses multimodal sentiment analysis to analyze user satisfaction of iPhone and Samsung products through online videos. The research reveals that the combination model integrating multiple data sources showed the most superior performance. Findings The findings also indicate that price is a crucial factor influencing user satisfaction, and users tend to exhibit more positive emotions when content with a product's price. The study highlights the importance of considering multiple factors when evaluating user satisfaction and provides valuable insights into the effectiveness of different data sources for sentiment analysis of product reviews.

• Journal of Digital Contents Society
• /
• v.18 no.3
• /
• pp.599-606
• /
• 2017

'Video Commerce' has grown significantly, and is in the era of so-called V-commerce 2.0. Based on this background, this study focused on the link and the possibility of creating synergy between V-commerce 2.0 content and MCN, and examined the linkage strategy considering its characteristics. In conclusion, first, V-Commerce has evolved into the age of 2.0, centered on the characteristics of content that are oriented towards fun and sympathy, beyond the 1.0 era. Second, V-commerce 2.0 content has the characteristic of replacing the sharing and recommendation based on the nature of SNS networks as promotion and purchase enhancement. Therefore, competitiveness as 'content' is relatively important before 'commerce'. Third, V-commerce 2.0 and MCN industry have a strong connection with each other in terms of securing core competitiveness and creating a new profit model. In order to create the synergy between V-Commerce 2.0 and MCN, we proposed the use of big data to reinforce V-Commerce 2.0 customized content competitiveness, building of storytelling marketing and branding, and enhancement of live performance and interactive communication.

• Journal of Broadcast Engineering
• /
• v.25 no.1
• /
• pp.25-35
• /
• 2020

Recently, the development of image processing technology, as well as hardware performance, has been continuing the research on 3D point processing technology that provides users with free viewing angle and stereoscopic effect in various fields. Point cloud technology, which is a type of representation of 3D point, has attracted attention in various fields because it can acquired/expressed point precisely. However, since Hundreds of thousands, millions of point are required to represent one 3D point cloud content, there is a disadvantage that a larger amount of storage space is required than a conventional 2D content. For this reason, the MPEG (Moving Picture Experts Group), an international standardization organization, is continuing to research how to efficiently compress, store, and transmit 3D point cloud content to users. In this paper, a V-PCC bitstream generated by a V-PCC (Video-based Point Cloud Compression) encoder proposed by the MPEG-I (Immersive) group is composed of an MPU (Media Processing Unit) defined by the MMT. In addition, by extending the signaling message defined in the MMT standard, a parameter for a segmented transmission method of the 3D point cloud content by area and quality parameters considering the characteristic of the 3D point cloud content, so that the quality parameters can be selectively determined according to the user's request. Finally, in this paper, we verify the result through design/implementation of the verification platform based on the proposed technology.

• Journal of KIISE:Computing Practices and Letters
• /
• v.8 no.6
• /
• pp.736-745
• /
• 2002

In network delivery of compressed video, packets may be lost if the channel is unreliable like Internet. Such losses tend to of cur in burst like continuous bit-stream error. In this paper, we propose an effective error-concealment approach to which an error resilient video encoding approach is applied against burst errors and which reduces a complexity of error concealment at the decoder using data hiding. To improve the performance of error concealment, a temporal and spatial error resilient video encoding approach at encoder is developed to be robust against burst errors. For spatial area of error concealment, block shuffling scheme is introduced to isolate erroneous blocks caused by packet losses. For temporal area of error concealment, we embed parity bits in content data for motion vectors between intra frames or continuous inter frames and recovery loss packet with it at decoder after transmission While error concealment is performed on error blocks of video data at decoder, it is computationally costly to interpolate error video block using neighboring information. So, in this paper, a set of feature are extracted at the encoder and embedded imperceptibly into the original media. If some part of the media data is damaged during transmission, the embedded features can be extracted and used for recovery of lost data with bi-direction interpolation. The use of data hiding leads to reduced complexity at the decoder. Experimental results suggest that our approach can achieve a reasonable quality for packet loss up to 30% over a wide range of video materials.