2nd Internatnional Conference on SMART Policing Convergence
Dongguk University, Seoul, Korea, OCT. 22. 2020
Application of Mobile Computer Vision for Dance Learning (Chungbuk National University)
Recently, with the worldwide popularity of K-POP, the number of people learning to dance through various video-sharing platforms, such as YouTube, has drastically increased. However, in the case of learning to dance individually, there is a problem that it is difficult to obtain objective feedback about whether the dance is correct. To solve this problem, several studies have been proposed that classify various dance gestures using 3-D joint information obtained from Kinect and evaluated the dance by analyzing the similarity of dance gestures between a teacher and student. However, these studies simply compare the similarity between the classified dance gestures without evaluation of the exact pose (i.e., position or angle of joints). Thus, it is difficult to use the results of existing studies for feedback on the exact pose required for dance learning. In addition, considering that these methods require special equipment, such as a 3-D camera, to obtain joint information, they have not been widely used for real-life dance learning.
In this talk, I will describe an application of mobile computer vision for dance learning. Specifically, I explain a new method to compare and evaluate the teacher and the student’s dance pose through the AI-based pose estimation obtained from the 2-D camera of a smartphone. I will also explain a pose evaluation method based on the affine transformation and OKS (Object Keypoint Similarity) that rectify different body ratios and compare the position and angle of the joints. In the end, I will demonstrate the effectiveness of the proposed solutions through extensive experiments that we carried out on a smartphone with real-world datasets. Experimental results showed that we could estimate the dance pose more accurately and provide feedback to the student in real-time.
Prof. Aziz Nasridinov is currently an Associate Professor of computer science with Chungbuk National University. He received the B.Sc. degree from the Tashkent University of Information Technologies in 2006 and the M.Sc. and Ph.D. degrees from Dongguk University in 2009 and 2012, respectively. Prof. Aziz Nasridinov has published over 20 papers in various high-ranked international journals and conferences. He has also served as a program committee member and co-organizer for numerous top-tier conferences, including ACM SAC, IEEE Big Data, IEEE Globecom and AAAI2021, and also served in the editorial board of several international journals. His research interests include traditional databases, big data analytics with machine learning, and computer vision.
Bibliometric investigation into the past and present of the computed tomography(CT) - related studies in the field of forensic science (Middle Tennessee State University)
Since its first development in the 1970’s, computed tomography (CT) has attracted much attention as a useful technique for a non-invasive diagnosis and examination in the fields of forensic science. Nowadays, CT is used not only to conduct a virtual autopsy in the medical examiner’s office but also to collect data for various research. This study aims to investigate the trend of CT-related forensic studies using the bibliometric method, which will provide researchers with a holistic insight into the application of CT in the field of forensic science.
2014 : Ph.D. in Anthropology; University of Tennessee, Knoxville (concentration: Forensic Anthropology)
2008 : M.A. in Anthroppology; Seoul National University (concentration: Physical Anthropology)
2003 : B.A. in Anthropology (2003); Seoul National University
General Conference Chair
Joon Tae Lim, Dongguk University
Yunsik Son, Dongguk University
Dongwook Kang, Dongguk University
Youngeun Song, Hoseo University
Junho Jeong, Kongju National University
Sun-Young Ihm, Dongguk University
Younoh Cho, Dongguk University
Jong Sup Lee, Dongguk University
Kyungseok Hu, Yonsei University
Yeonsoo Kim, Dongguk University
Jaehun Lee, Dongguk University
Hee Chang Jeon, Dongguk University