• 한국멀티미디어학회논문지
• /
• 제22권6호
• /
• 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.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권1호
• /
• pp.68-75
• /
• 1999

Machine recognition of faces from still and video images is emerging as an active research area spanning several disciplines such as image processing, pattern recognition, computer vision and neural networks. In addition, human face identification has numerous applications such as human interface based systems and real-time video systems of surveillance and security. In this paper, we propose an algorithm that can identify a particular individual face. We consider human face identification system in color space, which hasn't often considered in conventional in conventional methods. In order to make the algorithm insensitive to luminance, we convert the conventional RGB coordinates into normalized CIE coordinates. The normalized-CIE-based facial images are KL-transformed. The transformed data are used as the input of multi-layered neural network and the network are trained using error-backpropagation methods. Finally, we verify the system performance of the proposed algorithm by experiments.

• International journal of advanced smart convergence
• /
• 제11권3호
• /
• pp.64-71
• /
• 2022

Various types of attendance management systems are being introduced in a remote working environment and research on using face recognition is in progress. To ensure accurate worker's attendance, a face recognition-based attendance management system must analyze every frame of video, but face recognition is a heavy task, the number of the task should be minimized without affecting accuracy. In this paper, we proposed a search model using time interval variation to minimize the number of face recognition task of recorded videos for attendance management system. The proposed model performs face recognition by changing the interval of the frame identification time when there is no change in the attendance status for a certain period. When a change in the face recognition status occurs, it moves in the reverse direction and performs frame checks to more accurate attendance time checking. The implementation of proposed model performed at least 4.5 times faster than all frame identification and showed at least 97% accuracy.

• 한국항행학회논문지
• /
• 제12권6호
• /
• pp.691-698
• /
• 2008

본 논문에서는 비디오 검색을 위한 새로운 얼굴 검출 및 인식 방법을 제안한다. 인물 정함은 비디오 프레임에서 어떻게 얼굴을 정확하게 찾아내는가에 달려 있다. 얼굴 영역은 Adaboost 알고리즘으로 부스트된 viola-jones의 특징을 이용하여 비디오 프레임에서 검출한다. 얼굴 검출 후 조명 보정을 하고 PCA(Principal Component Analysis)로 특징점을 추출하고 SVM(Support Vector Machine)으로 사람의 신원을 분류한다. 실험 결과 제안한 방법이 정합율면에서 우수한 성능을 보였다.

• 공학교육연구
• /
• 제24권4호
• /
• pp.21-29
• /
• 2021

The objective of this research is to devise a BARS evaluation system as a performance evaluation plan for non-face-to-face capstone design and to verify the validity through the expert FGI as the remote education is highlighted as a new normal standard in the post corona epoch. The conclusion of this research is as follows. First, the non-face-to-face capstone design is a competency centered subject that allows you to develop the engineering and majoring knowledge and its function and attitude, and the achievement of program outcome is the objective competency, and the researcher proposes the BARS method evaluation, one of competency evaluation method, as a new performance evaluation plan. Second, for the evaluation of PO achievement of non-face-to-face capstone design, the researcher deduced 20 behavior identification standard(anchor) of BARS evaluation system, and developed the achievement standard per 4 levels. Third, as the evaluation tool of non-face-to-face capstone design, the presentation data(PPT), presentation video, product such as trial product(model), non-face-to-face class participation video, discussion participating video, team activity report, and result report for the evidential data of BARS evaluation were appeared as proper. Finally, the BARS evaluation plan of non-face-to-face capstone design would be efficiently made through the establishment of evaluation plan, the establishment of grading standard of BARS evaluation scale, the determination of evaluation subject and online BARS evaluation site.

• 한국정보통신학회논문지
• /
• 제25권1호
• /
• pp.63-68
• /
• 2021

최근 유튜브와 같이 영상 콘텐츠를 보거나 제작하는 것에 관한 관심이 급증하고 있습니다. 그러나 개인 정보 보호 기술이 없이 동영상을 제작하게 되면, 출연을 원하지 않는 사람들이 공개적으로 노출되어 개인 정보 보호권을 침해할 수 있습니다. 본 논문은 이러한 문제를 해결하기 위해 얼굴을 식별하여 특정한 얼굴만 화면에 나오고 그 외에 다른 얼굴들은 Gaussian blur filter를 이용하여 흐리게 하여서 초상권을 보호하는 기술을 제안합니다. 이 논문의 핵심은 실시간 비디오에서 인물의 초상권을 보호하기 위한 주요 기술인 얼굴 식별 기술의 정확도를 높이기 위한 노력입니다. 본 논문은 얼굴 식별의 정확도를 높이기 위하여 추적 알고리즘을 사용하였으며 실시간 비디오에 적용하기 위하여 알고리즘을 변경하였습니다. 이 논문에서는 추적 알고리즘이 있는 경우와 없는 경우를 비교하여 결과를 보여줍니다.

• 융합신호처리학회논문지
• /
• 제6권4호
• /
• pp.163-168
• /
• 2005

본 논문에서는 인물 기반의 비디오 요약 방법으로써 비디오 내 음성정보를 이용하여 화자 인식 기법을 통한 등장인물 중심의 요약 기법을 제안한다. 먼저, 얼굴 영역을 포함하는 장면을 중심으로 비디오로부터 배우의 대사에 해당하는 음성 정보를 분리하고, 화자 인식 기법을 수행하여 등장인물 별로 분류하였다. 화자인식 기법은 각 화자별로 MFCC(Mel Frequency Cepstrum Coefficient) 값을 추출하고 GMM(Gaussian Mixture Model)을 이용하여 분류한다. 본 논문에서는 4명의 등장인물에 대해 GMM을 학습시키고 4명 중 1명을 검출하는 실험을 통해 학습된 GMM 분류기가 실험 비디오에 대해 0.138 정도의 오분류율을 보임을 확인하였다.

• 정보보호학회논문지
• /
• 제21권1호
• /
• pp.85-91
• /
• 2011

영화, 드라마와 같은 비디오에서 등장하는 인물을 검출하고 인식하는 기술은 일반 사용자를 대상으로 PVR(personal video recorder), 모바일 폰과 멀티미디어 공급자를 위한 얼굴 영상의 데이터베이스 관리와 같은 응용분야에 이용될 수 있다. 본 논문에서는 비디오 등장인물을 자동으로 검색하는 얼굴검출 기법을 제안한다. 본 논문에서 제안한 얼굴검출 기법은 크게 세 단계로 구성하였다. 첫 번째, 정지 영상을 생성한 후 장면전환점 검출단계, 두 번째, 색상정보를 이용한 얼굴검출 단계, 마지막으로 얼굴영역을 이용한 얼굴 특징점 검출단계로 구성하였다. 실험결과 다양한 환경에서 성공적으로 얼굴을 검출하였으며, 기존의 색상정보를 이용한 얼굴 검출기법보다 성능이 개선된 것을 확인할 수 있었다.

• Journal of Information Processing Systems
• /
• 제5권2호
• /
• pp.41-68
• /
• 2009

Face recognition presents a challenging problem in the field of image analysis and computer vision, and as such has received a great deal of attention over the last few years because of its many applications in various domains. Face recognition techniques can be broadly divided into three categories based on the face data acquisition methodology: methods that operate on intensity images; those that deal with video sequences; and those that require other sensory data such as 3D information or infra-red imagery. In this paper, an overview of some of the well-known methods in each of these categories is provided and some of the benefits and drawbacks of the schemes mentioned therein are examined. Furthermore, a discussion outlining the incentive for using face recognition, the applications of this technology, and some of the difficulties plaguing current systems with regard to this task has also been provided. This paper also mentions some of the most recent algorithms developed for this purpose and attempts to give an idea of the state of the art of face recognition technology.

• Journal of Information Processing Systems
• /
• 제16권1호
• /
• pp.6-29
• /
• 2020

Biometrics identification using multiple modalities has attracted the attention of many researchers as it produces more robust and trustworthy results than single modality biometrics. In this paper, we present a novel multimodal recognition system that trains a deep learning network to automatically learn features after extracting multiple biometric modalities from a single data source, i.e., facial video clips. Utilizing different modalities, i.e., left ear, left profile face, frontal face, right profile face, and right ear, present in the facial video clips, we train supervised denoising auto-encoders to automatically extract robust and non-redundant features. The automatically learned features are then used to train modality specific sparse classifiers to perform the multimodal recognition. Moreover, the proposed technique has proven robust when some of the above modalities were missing during the testing. The proposed system has three main components that are responsible for detection, which consists of modality specific detectors to automatically detect images of different modalities present in facial video clips; feature selection, which uses supervised denoising sparse auto-encoders network to capture discriminative representations that are robust to the illumination and pose variations; and classification, which consists of a set of modality specific sparse representation classifiers for unimodal recognition, followed by score level fusion of the recognition results of the available modalities. Experiments conducted on the constrained facial video dataset (WVU) and the unconstrained facial video dataset (HONDA/UCSD), resulted in a 99.17% and 97.14% Rank-1 recognition rates, respectively. The multimodal recognition accuracy demonstrates the superiority and robustness of the proposed approach irrespective of the illumination, non-planar movement, and pose variations present in the video clips even in the situation of missing modalities.