• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.655-664
• /
• 2019

In this paper, we propose a novel way of combining multiple deep convolutional neural network (DCNN) architectures which work well for accurate video face identification by adopting a serial combination of 3D and 2D DCNNs. The proposed method first divides an input video sequence (to be recognized) into a number of sub-video sequences. The resulting sub-video sequences are used as input to the 3D DCNN so as to obtain the class-confidence scores for a given input video sequence by considering both temporal and spatial face feature characteristics of input video sequence. The class-confidence scores obtained from corresponding sub-video sequences is combined by forming our proposed class-confidence matrix. The resulting class-confidence matrix is then used as an input for learning 2D DCNN learning which is serially linked to 3D DCNN. Finally, fine-tuned, serially combined DCNN framework is applied for recognizing the identity present in a given test video sequence. To verify the effectiveness of our proposed method, extensive and comparative experiments have been conducted to evaluate our method on COX face databases with their standard face identification protocols. Experimental results showed that our method can achieve better or comparable identification rate compared to other state-of-the-art video FR methods.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.12 no.3
• /
• pp.171-190
• /
• 2008

With growing applications of wireless video streaming, an efficient video traffic model featuring modern high-compression techniques is more desirable than ever, because the wireless channel bandwidths are ever limited and time-varying. We propose a modeling and analysis method for video traffic by a class of stochastic processes, which we call '$GEO^Y/G/{\infty}$ input processes'. We model video traffic by $GEO^Y/G/{\infty}$ input process with gamma-distributed batch sizes Y and Weibull-like autocorrelation function. Using four real-encoded, full-length video traces including action movies, a drama, and an animation, we evaluate our modeling performance against existing model, transformed-M/G/${\infty}$ input process, which is one of most recently proposed video modeling methods in the literature. Our proposed $GEO^Y/G/{\infty}$ model is observed to consistently provide conservative performance predictions, in terms of packet loss ratio, within acceptable error at various traffic loads of interest in practical multimedia streaming systems, while the existing transformed-M/G/${\infty}$ fails. For real-time implementation of our model, we analyze G/D/1/K queueing systems with $GEO^Y/G/{\infty}$ input process to upper estimate the packet loss probabilities.

• Proceedings of the IEEK Conference
• /
• 2000.06d
• /
• pp.218-221
• /
• 2000

In this paper, we propose a new algorithm to detect human faces for controling a camera used in video conference. We model the distribution of skin color and set up the standard skin color in YIQ color space. An input video frame image is segmented into skin and non-skin segments by comparing the standard skin color and each pixels in the input video frame. Then, shape filler is applied to select face segments from skin segments. Our algorithm detects human faces in real time to control a camera to capture a human face with a proper size and position.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1996.06a
• /
• pp.153-158
• /
• 1996

Delay and jitter are critical factors in the real-time video services over ATM network. Mostly, delay and jitter problem are generated in input buffer when video are multiplexed. In this paper, we analyze delay and jitter of input buffer, and consider efficient control and flexible bandwidth allocation of video traffic. Also, we analyze decision of buffer size related maximum allowable delay.

• Journal of Information Technology Applications and Management
• /
• v.13 no.3
• /
• pp.75-87
• /
• 2006

In this paper. we consider the multi-view video technology. The technology, which is a field of 3-Dimensional video processing, enables the user to watch the various view-point of video. We expect that the technology will be applicable to a lot of video services. The economic effects of new technology are very important concern for the technology developer and the technology development policy makers. We, therefore. propose a general method for the economic repercussion effects of the multi-view video technology. The method is based on the expert opinion and input-output analysis. The results for the multi-view video technology are included.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.3
• /
• pp.1040-1057
• /
• 2014

In this paper, a new bit rate estimation scheme is proposed to determine the bit rate for each subclass in an MPEG-2 TS to H.264/AVC transcoder after dividing an input MPEG-2 TS sequence into several subclasses. Video format transcoding in conventional IPTV and Smart TV services is a time-consuming process since the input sequence should be fully transcoded several times with different bit-rates to decide the bit-rate suitable for a service. The proposed scheme can automatically decide the bit-rate for the transcoded video sequence in those services which can be stored on a video streaming server as small as possible without losing any subject quality loss. In the proposed scheme, an input sequence to the transcoder is sub-classified by hierarchical clustering using a parameter value extracted from each frame. The candidate frames of each subclass are used to estimate the bit rate using a statistical analysis and a mathematical model. Experimental results show that the proposed scheme reduces the bit rate by, on an average approximately 52% in low-complexity video and 6% in high-complexity video with negligible degradation in subjective quality.

• The Journal of Information Systems
• /
• v.32 no.2
• /
• pp.87-108
• /
• 2023

Purpose The main purpose of this study is to improve fake news detection performance by using video information to overcome the limitations of extant text- and image-oriented studies that do not reflect the latest news consumption trend. Design/methodology/approach This study collected video clips and related information including news scripts, speakers' facial expression, and video metadata from YouTube to develop fake news detection model. Based on the collected data, seven combinations of related information (i.e. scripts, video metadata, facial expression, scripts and video metadata, scripts and facial expression, and scripts, video metadata, and facial expression) were used as an input for taining and evaluation. The input data was analyzed using six models such as support vector machine and deep neural network. The area under the curve(AUC) was used to evaluate the performance of classification model. Findings The results showed that the ACU and accuracy values of three features combination (scripts, video metadata, and facial expression) were the highest in logistic regression, naïve bayes, and deep neural network models. This result implied that the fake news detection could be improved by using video information(video metadata and facial expression). Sample size of this study was relatively small. The generalizablity of the results would be enhanced with a larger sample size.

• Journal of Korea Society of Industrial Information Systems
• /
• v.8 no.2
• /
• pp.38-44
• /
• 2003

In this paper, we propose the method in which multimedia data simultaneously transfers to the main memory and the multimedia processor from the network interface card to improve bottleneck of system bus through analysis for architecture of video conference system and input/output model. The proposed method can reduce the number of system bus accesses, bus cycles, data transmission time and compression ratio of video data in the video conference system. We compared the performance between the proposed method and the conventional methods in the multi-party video conference systems. The simulation results showed that the proposed method was reduced the transmission time of multimedia data than the conventional method.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.14 no.1
• /
• pp.169-177
• /
• 2014

It is essential to support the image and video analysis technology such as motion detection since the DVR and NVR storage were adopted in the real time visual surveillance system. Especially the network camera would be popular as a video input device. The traditional CCTV that supports analog video data get be replaced by the network camera. In this paper, we present the design and implementation of video surveillance system that provides the real time motion detection by the video storage server. The mobile application also has been implemented in order to provides the retrieval functionality of image analysis results. We develop the video analysis server with open source library OpenCV and implement the daemon process for video input processing and real-time image analysis in our video surveillance system.

• The Journal of Bigdata
• /
• v.7 no.1
• /
• pp.99-111
• /
• 2022

A simulation technique can be used to analyze performance measures and support decision makings in manufacturing systems considering operational uncertainty and complexity. The simulation requires an input modeling procedure to reflect the target system's characteristics. However, data collection to build a simulation is quite limited when a target system includes manual productions with a lot of operational time such as construction equipment assembly lines. This study proposes a procedure for simulation input modeling using video data when it is difficult to collect enough input data to fit a probability distribution. We conducted a video-data analysis and specify input distributions for the simulation. Based on the proposed procedure, simulation experiments were conducted to evaluate key performance measures of the target system. We also expect that the proposed procedure may help simulation-based decision makings when obtaining input data for a simulation modeling is quite challenging.