Seung Hwan Kim, Ph.D.

LG AI Research, Vision Lab Leader,
Research Fellow

Skills

Computer Vision
Object Detection, Object Tracking, Image Classification, Face Recognition, Pose Estimation, Optical Character Recognition, Semantic Segmentation, 3D Reconstruction & Synthesis, Stereo Matching, Content-based Image/Video Retrieval
Image/Video Processing
Video Coding and Transmission, Digital Signal(Image/Video) Processing, Optical Character Recognition
Artificial Intelligence
Continual Learning, Explainable AI, Active Learning, Neural Rendering, Generative Model, Model Compression, Representation Learning, Deep Reinforcement Learning, Data Analytics for Prediction and Optimization
Project Manager
Tech. Strategy, Cloud System Architecture, System Engineer, System Operation

Education

PhD Electrical Engineering, Seoul National University Feb. 2004
MS Electrical Engineering, Seoul National University Feb. 1999
BS Electronics, Kyungpook National University Feb 1997

Work History

LG AI Research, Korea Dec. 2020 ~
- Vision Lab Leader, Research Fellow
LG Uplus, Korea Dec. 2018 to Nov. 2020
- Director, Head of AI Tech Unit
- Research Fellow, Head of Visual Intelligence
MindsLab, Korea March 2018 to Nov. 2018
- Vice President, Head of Video Analytics R&D Center
SK Telecom, Korea July 2006 to Feb. 2018
- Lead Video Surveillance and Analytics Engineer
- Senior Visual Intelligence Research Engineer
Samsung Electronics, Korea March 2004 to June 2006
- Senior Research and Development Engineer

Selected Project Experiences

3D Face Reconstruction based on Neural Rendering Dev. 2020 ~
Anomaly/Novelty Detection based on Generative model for Vision Inspection Dev. 2020 ~
Continual Learning for Vision Inspection Dev. 2020 ~
Explainable AI for Vision Inspection Dev. 2020 ~
Few Shot Video-to-Video Synthesis Technique using Generative Adversarial Network May 2020 to Nov. 2020
Scene Analysis using Multimodal Information for TV show and Movie Jan. 2020 to Oct. 2020
Video Content Recommendation System based on Hierarchical Recurrent Neural Network(H-RNN) Jan. 2020 to Oct. 2020
Multi-label Text Classification using BERT Jan. 2020 to Nov. 2020
AI Chatbot for Contact Center and Sales based-on BERT April 2019 to Nov. 2020
CNN-based Object Detection and Tracking for Video Surveillance: Intrusion Detection,
Loitering Detection, Fire Detection and Face Recognition Dec. 2018 to July 2020
YOLO-based Pet Detection Jan 2020 to April 2020
Multiple Human/Car Object Detection and Tracking at a distance based-on YOLOv3 for Drone Service Sept. 2019 to March 2020
Content-based Image/Video Retrieval for Copyright Protection based-on Deep Neural Network June 2018 to Dec.2018
Machine Learning based Process Optimization of Cold Mill Manufacturing for Steel Making Company May 2018 to Oct. 2018
Optical Character Recognition based-on LSTM-RNN and Currency Recognition based-on InceptionNet April 2018 to Sept. 2018
Video Cloud Platform and Intelligent Video Analytics for Video Surveillance Sept. 2013 to Feb. 2018
ASR and Natural Language Processing for Automatic English Speaking and Writing Assessment July 2010 to Aug. 2013
Video Communication system and its application for CDMA EVDO Network July 2006 to Dec. 2007
MPEG-4 and H.264 Video Codec for Mobile Application Processor Jan. 2005 to June 2006
MPEG-2, MPEG-4 Video Codec for High Definition System Jan. 2004 to Dec. 2004
Measurement of Video Subjective Quality in IS-95C Network June 2000 to May 2001
Development of 2k/8k FFT Processor for DVB-T May 2000 to Feb. 2001
MPEG-21 Standardization Dec. 1999 to Nov. 2000

Selected Publications

International Journal
Seung Hwan Kim, Chang-Su Kim, and Sang-Uk Lee, “An efficient motion compensation algorithm based on double reference frame method,: Signal Processing: Image Communication, vol. 17, no. 8, pp. 635-649, Sept. 2002.
Dong-il Chang, Seung Hwan Kim, and C. W. Lee, “Fast motion search for warping prediction with classification,” in Electronics letters, vol. 34, no. 10, pp 973-974, May., 1998.

International Conference
Seung Hwan Kim, and Yong Kwan Kim, and Sang-Uk Lee, “Error-resilient video coding using adaptive weighted multiple reference frame method,” Proc. VCIP-2004, Jan. 2004.
Seung Hwan Kim, and Sang-Uk Lee, “Error-resilient video coding using control grid interpolation for H.264,” IEEE midwest symposium on circuits and systems, Dec. 2003.
Seung Hwan Kim, and Yong Kwan Kim, and Sang-Uk Lee, “Adaptive multiple reference frame based scalable video coding algorithm,” Proc. ICIP-2002, vol. 2, pp. 33-36, Sept. 2002
Seung Hwan Kim, Yong Kwan Kim, Kyeong Ho Yang, and Sang-Uk Lee, “Multiple reference frame based scalable video coding for low-delay internet transmission,” Proc. International workshop on very low bitrate video coding, pp. 181-186, Oct. 2001.
Seung Hwan Kim, Chang-Su Kim, and Sang-Uk Lee, “Efficient motion compensation algorithm based on second-order prediction,” Proc. ICIP-2000, vol. 2, pp. 875-878, Sept. 2000.
Seung Hwan Kim, Dong-il Chang, Choong Woong Lee, and Snag-Uk Lee, “Complexity reduction method for overlapped block motion compensation based on spatio-temporal correlation,” Proc. ISCAS-1999, vol. 4, pp. 211-214, May 1999.
Dong-il Chang, Seung Hwan Kim, Jeong-Kwon Kim, and Choong Woong Lee, “A fast motion search for warping prediction based on node classification,” Proc. GLOBECOM-1998, vol. 2, pp. 862-867, Nov. 1998.
Seung Hwan Kim, Dong-il Chang, Jeong-Kwon Kim, and Choong Woong Lee, “Fast motion search for overlapped block motion compensation methods based on block-adaptive classification,” Proc. IEEE international symposium on intelligent signal processing and communication systems, vol. 1, pp. 63-67, Nov. 1998.
Jeong-Kwon Kim, Dong-il Chang, Seung Hwan Kim, and Choong Woong Lee, “Fast classified de-blocking algorithm and derivation of loop filter similar to H.263+’s,” ” Proc. IEEE international symposium on intelligent signal processing and communication systems, vol. 1, pp. 58-62, Nov. 1998.
Dong-il Chang, Seung Hwan Kim, Jeong-Kwon Kim, and Choong Woong Lee, “A fast motion search for warping prediction based on node classification,” Proc. HDTV workshop, pp. 120-129, Oct. 1998.
Hyo June Kim, June Seok song, Seung Hwan Kim, Jong Hyo Kim, and Choong Woong Lee, “Region adaptive JPEG encoding scheme for medical image,” Proc. SPIE medical imaging: image display, vol. 3335, pp. 595-599, Feb. 1998.

Selected International Patents

Method for detecting a grammatical error and apparatus thereof, Unites States Patent, No. 9,600,469.
Method for detecting grammar error and apparatus thereof, Unites States Patent, No. 9,575,955
Example based error detection system and method for estimating writing automatically, , China Patent, No. ZL201280034880.3
Method for detecting grammar error and apparatus thereof, Unites States Patent, No. 9,575,955
Error Detection Apparatus and Method Based on Shallow Parser for Estimating Writing Automatically, Unites States Patent, No. 9,336,199
System and method for widget service based on smart card, and smart card applied on the same, China Patent, No. ZL201080017335.4
System and method for browsing based on smart card, and smart card applied to the same, China Patent, No. ZL201080003712.9
System and method for playing contents based on smart card, and smart card applied to the same, China Patent, No. ZL201080019572.4
System for providing an advertisement service based on smart card and method thereof, Japan Patent, No 5363574

Seung Hwan Kim, Ph.D.

LG AI Research, Vision Lab Leader, Research Fellow

Skills

Education

Work History

Selected Project Experiences

Selected Publications

Selected International Patents

sh.kim@lgresearch.ai

LG AI Research, Vision Lab Leader,
Research Fellow