Current Projects

1. Complementary Deep Learning Model for Pixel-level Correspondence/Segmentation and Image-level Retrieval
   • Period : Mar.2018 ~ Feb.2021
   • Organization : Funded by National Research Foundation (NRF) of Korea, Korea (100M KRW/year)
   • Role : Principal Investigator (PI), Chungnam National University, Korea
2. A study on highly efficient and advanced visual correspondence algorithms for big visual data
   • Period : Nov. 2015 ~ Oct. 2018
   • Organization : Funded by National Research Foundation (NRF) of Korea, Korea (51M KRW/year)
   • Role : Principal Investigator (PI), Chungnam National University, Korea
3. High Quality 2D-to-Multiview Contents Generation from Large-scale RGB+D Database
   • Period : July. 2015 ~ Aug. 2017
   • Organization : Funded by Institute for Information & Communications Technology Promotion (IITP), Korea (80M KRW/year)
   • Role : Co-Principal Investigator (Co-PI), Chungnam National University, Korea

Previous Projects

1. Visual Modeling and Analytics of Dynamic Environments for the Mass
   • Period : Apr. 2014 ~ Feb. 2015
   • Organization : Funded by Advance Digital Science Center (ADSC), Singapore
   • Role : Research scientist, Advance Digital Science Center (ADSC), Singapore
   • Project Details [link]
     • eveloped a fast inference machine (e.g. global optimization) that is suitable in handling and labeling the large-scale, high-dimensional visual data (Big visual data analysis).
2. Remote Reality for Immersive Communications and Games
   • Period : Jul. 2010 ~ Mar. 2014
   • Organization : Funded by Advance Digital Science Center (ADSC), Singapore
   • Role : Researcher, and later Research scientist, Advance Digital Science Center (ADSC), Singapore
   • Project Details [link]
     • Developed novel edge-preserving filters as generic computing tools for image processing and computer vision applications. The overview of this work was presented in a half-day tutorial in ICIP 2013.[link]
     • Developed fast and high-quality image/video editing and enhancement tools.
     • Developed fast discrete labeling algorithms for Nearest Neighbor Fields (NNF) search and depth/optical flow estimation.
     • Developed reliable video based rendering (VBR) algorithms for freeview video generation.
     • Developed novel methods for depth enhancement and coding using active depth sensor such as Time-of-Flight (ToF) camera and Kinect.
     • Developed a tele-conferencing system using MS Kinect (with image based rendering and object segmentation).
     • Developed real-time image processing algorithms on GPUs with interns.
     • Collaborated with the embedded system group (ADSC) for FPGA implementation of computer vision algorithms using high level synthesis tools.
3. 3D Display Processing Techniques
   • Period : Jun. 2009 ~ Jun. 2010
   • Organization : Funded by Mitsubishi Electric Research Laboratories (MERL), US
   • Role : Post-doctoral researcher, Mitsubishi Electric Research Laboratories (MERL), US
   • Project Details [link]
     • Developed a prototype of 3D video system (3DTV)
     • Developed efficient and accurate depth estimation and occlusion handling methods for 3D video processing
     • Developed 3D video processing method with temporal coherence

Current Projects

1. Complementary Deep Learning Model for Pixel-level Correspondence/Segmentation and Image-level Retrieval
   • Period : Mar.2018 ~ Feb.2021
   • Organization : Funded by National Research Foundation (NRF) of Korea, Korea (100M KRW/year)
   • Role : Principal Investigator (PI), Chungnam National University, Korea
2. A study on highly efficient and advanced visual correspondence algorithms for big visual data
   • Period : Nov. 2015 ~ Oct. 2018
   • Organization : Funded by National Research Foundation (NRF) of Korea, Korea (51M KRW/year)
   • Role : Principal Investigator (PI), Chungnam National University, Korea
3. High Quality 2D-to-Multiview Contents Generation from Large-scale RGB+D Database
   • Period : July. 2015 ~ Aug. 2017
   • Organization : Funded by Institute for Information & Communications Technology Promotion (IITP), Korea (80M KRW/year)
   • Role : Co-Principal Investigator (Co-PI), Chungnam National University, Korea

Previous Projects

1. Visual Modeling and Analytics of Dynamic Environments for the Mass
   • Period : Apr. 2014 ~ Feb. 2015
   • Organization : Funded by Advance Digital Science Center (ADSC), Singapore
   • Role : Research scientist, Advance Digital Science Center (ADSC), Singapore
   • Project Details [link]
     • eveloped a fast inference machine (e.g. global optimization) that is suitable in handling and labeling the large-scale, high-dimensional visual data (Big visual data analysis).
2. Remote Reality for Immersive Communications and Games
   • Period : Jul. 2010 ~ Mar. 2014
   • Organization : Funded by Advance Digital Science Center (ADSC), Singapore
   • Role : Researcher, and later Research scientist, Advance Digital Science Center (ADSC), Singapore
   • Project Details [link]
     • Developed novel edge-preserving filters as generic computing tools for image processing and computer vision applications. The overview of this work was presented in a half-day tutorial in ICIP 2013.[link]
     • Developed fast and high-quality image/video editing and enhancement tools.
     • Developed fast discrete labeling algorithms for Nearest Neighbor Fields (NNF) search and depth/optical flow estimation.
     • Developed reliable video based rendering (VBR) algorithms for freeview video generation.
     • Developed novel methods for depth enhancement and coding using active depth sensor such as Time-of-Flight (ToF) camera and Kinect.
     • Developed a tele-conferencing system using MS Kinect (with image based rendering and object segmentation).
     • Developed real-time image processing algorithms on GPUs with interns.
     • Collaborated with the embedded system group (ADSC) for FPGA implementation of computer vision algorithms using high level synthesis tools.
3. 3D Display Processing Techniques
   • Period : Jun. 2009 ~ Jun. 2010
   • Organization : Funded by Mitsubishi Electric Research Laboratories (MERL), US
   • Role : Post-doctoral researcher, Mitsubishi Electric Research Laboratories (MERL), US
   • Project Details [link]
     • Developed a prototype of 3D video system (3DTV)
     • Developed efficient and accurate depth estimation and occlusion handling methods for 3D video processing
     • Developed 3D video processing method with temporal coherence