Publications & Awards

Awards

  • CVPR 2021 NAS competition unseen data track 3위 수상 [관련 기사]

  • NeurIPS 2019 AutoDL competition Computer Vision track 2위 수상 [관련 기사]

  • NeurIPS 2019 AutoDL competition NLP track on-line phase 2위

Publications

  • Do-Guk Kim and Heung-Chang Lee, "Proxyless Neural Architecture Adaptation at Once," IEEE Access (September 2022), DOI: 10.1109/ACCESS.2022.3206765

  • Jongheon Jeong, Sejun Park, Minkyu Kim, Heung-Chang Lee, Do-Guk Kim, Jinwoo Shin, "SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness," Neural Information Processing Systems (NeurIPS 2021), 6-14 Dec 2021

  • Do-Guk Kim* and Heung-Chang Lee*, “Differentiable Neural Architecture Transformation for Reproducible Architecture Improvement,” CVPR 2020 Workshop on Neural Architecture Search and Beyond for Representation Learning (CVPRNAS 2020), 16-18 June 2020

  • Heung-Chang Lee*, Do-Guk Kim*, and Bohyung Han, “Efficient Decoupled Neural Architecture Search by Structure and Operation Sampling,” International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2020), 4-8 May 2020

(*: Equal contribution & Corresponding author)

  • Do-Guk Kim and Bong-Gyun Ko, “Investment Universe Construction based on the Theme Keyword Search,” IEEE Access (September 2019), DOI: 10.1109/ACCESS.2019.2939414

  • Do-Guk Kim, Jong-Uk Hou, and Heung-Kyu Lee, “Learning Deep Features for Source Color Laser Printer Identification based on Cascaded Learning,” Neurocomputing (July 2019), DOI: 10.1016/j.neucom.2019.07.084

  • Jin-Seok Park, Hyeon-Ki Kim, Do-Guk Kim, In-Jae Yu, and Heung-Kyu Lee, “Paired mini-batch training: A new deep network training for image forensics and steganalysis,” Signal Processing: Image Communication, Vol. 67, (September 2018), pp. 132-139.

  • Hyeon-Ki Kim, Jin-Seok Park, Do-Guk Kim, and Heung-Kyu Lee, “Two-stream neural networks to detect manipulation of JPEG compressed images,” Electronics Letters, Vol. 54, No. 6 (22 March 2018), pp. 354-355.

  • In-Jae Yu, Do-Guk Kim, Jin-Seok Park, Jong-Uk Hou, Sunghee Choi, and Heung-Kyu Lee, “Identifying Photorealistic Computer Graphics using Convolutional Neural Networks,” IEEE Int. Conf. on Image Processing (ICIP 2017), 17-20 September 2017, Beijing, China

  • Jin-Seok Park, Dai-Kyung Hyun, Jong-Uk Hou, Do-Guk Kim, and Heung-Kyu Lee, “Detecting Digital Image Forgery in Near-Infrared Image of CCTV,” Multimedia Tools and Applications, Vol. 76, No. 14 (July 2017), DOI:10.1007/s11042-016-3871-7

  • Do-Guk Kim and Heung-Kyu Lee, “Color Laser Printer Identification Using Halftone Texture Fingerprint,” Electronics Letters, Vol. 51, No. 13 (25 June 2015), pp. 981-983.

  • Do-Guk Kim and Heung-Kyu Lee, “Color Laser Printer Identification Using Photographed Halftone Images,” EURASIP, 22th European Signal Processing Conf. (EUSIPCO 2014), 1-5 September 2014, Lisbon, Portugal, pp. 795-799.

  • Jong-Uk Hou, Jin-Seok Park, Do-Guk Kim, Seung-Hun Nam, and Heung-Kyu Lee, “Robust Video Watermarking for MPEG Compression and DA-AD Conversion,” ACM The 1st Int. Workshop on Information Hiding and Its Criteria for Evaluation (IWIHC 2014), 3 June 2014, Kyoto, Japan, pp. 2-8.

Patents

  • Heung-Kyu Lee, Jin-Seok Park, Hyeon-Ki Kim, Do-Guk Kim, In-Jae Yu, Seung-Hun Nam, and Tae-Yeon Kim, “Method and apparatus for training neural network,” Registration number: 10-2223070-0000

  • Heung-Kyu Lee, Jong-Uk Hou, Dong-Kyu Kim, Hyun-Ji Song, and Do-Guk Kim, “Method for processing secure image through visual cryptography based on temporal domain and apparatus therefor,” Registration number: 10-1879954-0000

  • Heung-Kyu Lee and Do-Guk Kim, “Color Laser Printer Identification Method Using Photographed Halftone Images and Apparatus therefor,” Registration number: 10-1629138-0000