Home News Research Projects Teaching
profile photo

Changjae Oh

I am a Lecturer (Assistant Professor) at the School of Electrical Engineering and Computer Science (EECS), Queen Mary University of London (QMUL) . I was a postdoctoral researcher at QMUL (May 2018-Aug 2019). I received the PhD degree in Electrical and Electronic Engineering from Yonsei University, South Korea.

Email  /  GitHub  /  Google Scholar  /  CV

Prospective Students

I'm looking for prospective students who are interested in computer vision or vision-based machine perception for robot manipulation. Please feel free to contact me. You can find information on the QMUL EECS PhD opportunities here, with some scholarships such as:

  • China Scholarship Council (CSC) Scholarships
  • Pakistan HEC-QMUL PhD Scholarships
  • QMUL scholarships database
  • News

  • (Oct. 2021) A paper on video prediction was accepted to BMVC.
  • (Sep. 2021) A paper on adversarial image filters was accepted to TIP.
  • (Jun. 2021) A paper on policy learning from a static image was accepted to IROS.
  • (Jun. 2021) A paper on safe human-to-robot handover was accepted to ROMAN.
  • (Jun. 2021) A paper on visual-tactile perception was accepted to CASE.
  • (Sep. 2020) Co-organised 2020 Intelligent Sensing Summer School
  • (Mar. 2020) A paper on mapless navigation was accepted to TCSVT.
  • (Jan. 2020) A paper on adversarial image enhancement filter was accepted to ICASSP.
  • (Jan. 2020) A paper on stereo dehazing was accepted to IJCV.
  • (Sep. 2019) I've joined as a faculty (Lecturer) at Queen Mary University of London.
  • Research

    I'm interested in image processing, computer vision, and vision-based machine perception.

    [J11] Semantically Adversarial Learnable Filters


    Ali Shahin Shamsabadi, Changjae Oh, Andrea Cavallaro
    IEEE Trans. Image Process. (TIP), vol. 30, pp. 8075-8087, Sep. 2021
    [arxiv] [code]

    [C21] OHPL: One-shot Hand-eye Policy Learner


    Changjae Oh, Yik Lung Pang, Andrea Cavallaro
    in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), Sep. 2021.
    [paper] [webpage]

    [C20] Towards safe human-to-robot handovers of unknown containers


    Yik Lung Pang, Alessio Xompero, Changjae Oh, Andrea Cavallaro
    in Proc. IEEE Int. Conf. Robot Human Interactive Commun. (ROMAN), Aug. 2021.
    [paper] [code] [webpage]

    [C19] Multi-modal robotic visual-tactile localisation and detection of surface cracks


    Francesca Palermo, Liz Katherine Rincon Ardila, Changjae Oh, Kaspar Althoefer, Stefan Poslad, Gentiane Venture, Ildar Farkhatdinov
    in Proc. IEEE Int. Conf. Autom., Sci. Eng. (CASE), 2021. (accepted)
    [paper]

    [J10] View-Action Representation Learning for Active First-Person Vision


    Changjae Oh and Andrea Cavallaro
    IEEE Trans. Circuits Syst. Video Technol. (TCSVT), vol. 32, no. 2, pp. 480-491, Feb. 2021.
    [paper]

    [J9] Simultaneous Deep Stereo Matching and Dehazing with Feature Attention


    Taeyong Song, Youngjung Kim, Changjae Oh, Hyunsung Jang, Namkoo Ha, and Kwanghoon Sohn
    Int. J. Comput. Vis. (IJCV), vol. 128, pp. 799-817, Jan. 2020.
    [paper]

    [C18] Edgefool: An Adversarial Image Enhancement Filter


    Ali Shahin Shamsabadi, Changjae Oh, and Andrea Cavallaro
    in Proc. IEEE Int. Conf. Acoust., Speech, Signal Process. (ICASSP), May 2020.
    [paper] [code]

    [J8] OCEAN: Object-Centric Arranging Network for Self-supervised Visual Representations Learning


    Changjae Oh, Bumsub Ham, Hansung Kim, Adrian Hilton, and Kwanghoon Sohn
    Expert Systems with Applications. (ESWA), vol. 125, pp. 281-292 Jul. 2019.
    [paper]

    [C17] Learning Action Representations for Self-supervised Visual Exploration


    Changjae Oh and Andrea Cavallaro
    in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), May 2019.
    [paper]

    [C16] Deep Network for Simultaneous Stereo Matching and Dehazing


    Taeyong Song, Youngjung Kim, Changjae Oh, and Kwanghoon Sohn
    in Proc. British Machine Vis. Conf. (BMVC), Sep. 2018. (Oral Presentation) (Best Science Paper Honourable Mention)
    [paper]

    [C15] Multi-task Self-supervised Visual Representation Learning for Monocular Road Segmentation


    Jaehoon Cho, Youngjung Kim, Hyungjoo Jung, Changjae Oh, Jaesung Youn, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Multimedia and Expo. (ICME), Jul. 2018. (Oral Presentation)
    [paper]

    [C14] Depth Prediction from a Single Image with Conditional Adversarial Networks


    Hyungjoo Jung, Youngjung Kim, Dongbo Min, Changjae Oh, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Image Process. (ICIP), Sep. 2017.
    [paper]

    [J7] Robust Interactive Image Segmentation using Structure-aware Labeling


    Changjae Oh, Bumsub Ham, and Kwanghoon Sohn
    Expert Systems with Applications. (ESWA), vol. 79, pp. 90-100, Aug. 2017.
    [paper]

    [J6] Structure selective depth super-resolution for RGB-D cameras


    Youngjung Kim, Bumsub Ham, Changjae Oh, and Kwanghoon Sohn
    IEEE Trans. Image Process. (TIP), vol. 26, pp. 4079-4091, Aug. 2017.
    [paper]

    [J5] Non-Parametric Human Segmentation Using Support Vector Machine


    Kyuwon Kim, Changjae Oh, and Kwanghoon Sohn
    IEEE Trans. Consumer Electronics (TCE), vol. 63, pp.93-100, May 2017.
    [paper]

    [J4] Personness estimation for real-time human detection on mobile devices


    Kyuwon Kim, Changjae Oh, and Kwanghoon Sohn
    Expert Systems with Applications (ESWA), vol. 72, pp. 130-138, Apr. 2017.
    [paper]

    [C13] Point-cut: Interactive Image Segmentation using Point Supervision


    Changjae Oh, Bumsub Ham, and Kwanghoon Sohn
    in Proc. Asian Conf. Comput. Vis. (ACCV), Nov. 2016.
    [paper]

    [C12] Edge-Aware Image Smoothing using Commute Time Distances


    Youngjung Kim, Changjae Oh, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Image Process. (ICIP), Sep. 2016.
    [paper]

    [C11] Depth Extraction from a Single Image Based on Block-Matching and Robust Regression


    Hyungjoo Jung, Changjae Oh, Youngjung Kim, and Kwanghoon Sohn
    in Proc. Electronic Imaging (EI), Feb. 2016.
    [paper]

    [C10] Non-Parametric Human Segmentation Using Support Vector Machine


    Kyuwon Kim, Changjae Oh, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Consumer Electronics (ICCE), Jan. 2016.

    [J3] Depth Analogy: Data-driven Approach for Single Image Depth Estimation using Gradient Samples


    Sunghwan Choi, Dongbo Min, Bumsub Ham, Youngjung Kim, Changjae Oh, and Kwanghoon Sohn
    IEEE Trans. Image Process. (TIP), vol. 24, no. 12, pp. 5953-5966, Dec. 2015.
    [paper]

    [C9] Segment-based Free Space Estimation using Plane Normal Vector in Disparity Space


    Jeonghyun Seo, Changjae Oh, and Kwanghoon Sohn
    in Proc. Int. Conf. Connected Vehicles & Expo, Oct. 2015.
    [paper]

    [C8] Sparse Edit Propagation for High Resolution Image using Support Vector Machines


    Changjae Oh, Seungchul Ryu, Youngjung Kim, Taewoong Park, Jihyun Kim, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Image Process. (ICIP), Sep. 2015.
    [paper]

    [C7] A Majorize-minimize Approach for High-Quality Depth Upsampling


    Youngjung Kim, Sunghwan Choi, Changjae Oh, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Image Process. (ICIP), Sep. 2015.
    [paper]

    [C6] Structure-Aware Depth Super-Resolution Using Gaussian Mixture Model


    Sunok Kim, Changjae Oh, Youngjung Kim, and Kwanghoon Sohn
    in Proc. SPIE Electronic Imaging (EI), Feb. 2015.

    [J2] Visual Fatigue Relaxation for Stereoscopic Video via Nonlinear Disparity Remapping


    Changjae Oh, Bumsub Ham, Sunghwan Choi, and Kwanghoon Sohn
    IEEE Trans. Broadcast. (TB), vol. 61, no. 2, pp. 142-153, Jun. 2015.
    [paper]

    [J1] Probability-Based Rendering for View Synthesis


    Bumsub Ham, Dongbo Min, Changjae Oh, Minh N. Do, and Kwanghoon Sohn
    IEEE Trans. Image Process. (TIP), vol. 23, no. 2, pp. 870-884, Feb. 2014.
    [paper]

    [C5] Visual Fatigue Prediction and Its Visualization


    Changjae Oh, Bumsub Ham, and Kwanghoon Sohn
    in Proc. Global 3D Tech Forum, Oct. 2013. (Best Paper Award)

    [C4] Disparity Search Range Estimation Based on Dense Stereo Matching


    Ruei-Hung Li, Bumsub Ham, Changjae Oh, and Kwanghoon Sohn
    in Proc. IEEE Int. Conf. Industrial Electronics and Applications (ICIEA), Jun. 2013. (Best Paper Award)
    [paper]

    [C3] Probabilistic Correspondence Matching using Random Walk with Restart


    Changjae Oh, Bumsub Ham, and Kwanghoon Sohn
    in British Machine Vis. Conf. (BMVC), Sep. 2012.
    [paper]

    [C2] Hybrid Approach for Accurate Depth Acquisition with Structured Light and Stereo Camera


    Sunghwan Choi, Bumsub Ham, Changjae Oh, Hyon-gon Choo, Jinwoong Kim, and Kwanghoon Sohn
    in Proc. IEEE Int. Symposium on Broadband Multimedia Systems and Broadcasting (BMSB), Jun. 2012.
    [paper]

    [C1] Joint Radiometric Calibration and Multi-view Matching with Ground Control Points


    Changjae Oh, Bumsub Ham, Hyon-gon Choo, Jinwoong Kim, Kwanghoon Sohn
    in Proc. IEEE Int. Workshop on Advanced Image Technology (IWAIT), Jan. 2012.

    Projects

    [May. 2018 - Aug. 2019] Nuclear Centre for Nuclear Robotics (NCNR)


    Postdoctoral research assistant
    The Engineering and Physical Sciences Research Council (EPSRC), UK
    Vision-based robot navigation for unseen environment

    Teaching

    [2020/2021] Computer Vision

    [2020/2021] Image and Video Processing

    [2020] Interactive Media Design and Production

    [2019] Advanced Transform Methods


    Design and source code from Jon Barron's website