Digital Image Media Lab is a research lab based in Yonsei University, Seoul, South Korea. Since the lab was founded in 1996 under supervision of Prof. Kwanghoon Sohn, we have been conducting research on diverse topics in image processing and computer vision area.

Some of the selected topics we are interested and currently working on are:

  • 3D Video Processing
  • Deep Learning for 3D Video Processing
    Multi-view Video Processing
    Image Enhancement / Editing

  • Intelligent Computer Vision
  • Deep Learning for Vision
    ADAS(advanced driver assistance systems)
    Multispectral Imaging / Feature Extraction


Recent activities

  • CIC26 2018, Vancouver, Canada: Y. Cho, "Reversible Colour Appearance Scales for Describing Saturation, Vividness, Blackness, and Whiteness for Image Enhancement"
  • ACMMM Workshop 2018, Seoul, Korea: J. Lee, "Audio-Visual Attention Networks for Emotion Recognition"
  • ACMMM Workshop 2018, Seoul, Korea: J. Park, "Learning to Detect, Associate, and Recognize Human Actions and Surrounding Scenes in Untrimmed Videos"
  • ECCV 2018, Munich, Germany: S. Jeon, "PARN : Pyramidal Affine Regression Networks for Dense Semnatic Correspondence Estimation"
  • BMVC 2018, Newcastle upon Tyne, UK: T. Song, "Deep Network for Simultaneous Stereo Matching and Dehazing"
  • ICME 2018, Sandiego, US: J. Cho, "Multi-task Self-supervised Visual Representation Learning for Monocular Road Segmentation"
  • CVPR Workshop 2018, Salt Lake City, US: S. Choi, "Learning Descriptor, Confidence, and Depth Estimation in Multi-view Stereo"
  • ICRA 2018, Brisbane, Australia: K. Park, "High-precision Depth Estimation with the 3D LiDAR and Stereo Fusion"
  • ICASSP 2018, Calgary, Canada: J. Lee, "Spatiotemporal Attention Based Deep Neural Networks for Emotion Recognition"
  • ICCV 2017, Venice, Italy: S. Kim, "DCTM: Discrete-Continuous Transform Matching for Semantic Flow"
  • ICIP 2017, Beijing, China: K. Park, "Pedestrian Proposal Generation using Depth-aware Scale Estimation"
  • ICIP 2017, Beijing, China: S. Kim, "Deep Stereo Confidence Prediction for Depth Estimation"
  • ICIP 2017, Beijing, China: H. Jung, "Depth Prediction from a Single Image with Conditional Adversarial Networks"
  • ICIP 2017, Beijing, China: S. Choi, "Multi-spectral Human Co-segmentation via Joint Convolutional Neural Networks"
  • ICIP 2017, Beijing, China: J. Lee, "Automatic 2D-to-3D Conversion using Multi-scale Deep Neural Network"
  • ICIP 2017, Beijing, China: S. Joung, "Unsupervised Stereo Matching using Correspondence Consistency"
  • ICIP 2017, Beijing, China: S. Jeon, "Convolutional Feature Pyramid Fusion via Attention Network"
  • ICIP 2017, Beijing, China: S. Jeong, "Convolutional Cost Aggregation for Robust Stereo Matching"
  • CVPR 2017, Hawaii, US: Y. Kim, "Deeply Aggregated Alternating Minimization for Image Resotration"
  • CVPR 2017, Hawaii, US: K. Kim, "FCSS: Fully Convolutional Self-Similarity for Dense Semantic Correspondence"
  • ICCE 2017, Las Vegas, US: K. Kim, "Lazy Dragging: Effortless Bounding-box Drawing for Touch-screen Devices"
  • APSIPA 2016, Jeju, KOR: K. Park, "Homography Flow for Dense Correspondences"
  • ICPR 2016, Cancun, MEX: H. Choi, "Multi-spectral Pedestrian Detection Based on Accumulated Object Proposal with Fully Convolution Network"
  • ACCV 2016, Taipei, TPE: C. Oh, "Point-cut: Interactive Image Segmentation using One-point Supervision"
  • ECCV 2016, Amsterdam, NLD: S. Kim, "Deep Self-Convolutional Activations Descriptor for Dense Cross-Modal Correspondence"
  • ECCV 2016, Amsterdam, NLD: S. Kim, "Unified Depth Prediction and Intrinsic Image Decomposition from a Single Image via Joint Convolutional Neural Fields"
  • ICIP 2016, Arizona, US: S. Kim, "ANCC FLOW: Adaptive Normalized Cross-Correlation with Evolving Guidance Aggregation for Dense Correspondence Estimation"
  • EI 2016, San Francisco, US: S. Choi, "Multi-level segment based dense correspondence: an affine transformation approach"
  • EI 2016, San Francisco, US: H. Jeong, "Depth Extraction from a Single Image Based on Block-Matching and Robust Regression"
  • ICCE 2016, Las Vegas, US: K. Kim, "Non-Parametric Human Segmentation Using Support Vector Machine"


Introduction Research People Lecture Publications Photo Contact