Research

My research interests include

  • Computer Vision
  • Medical Image Analysis
  • Machine Learning, Deep Learning
  • 3D Human Pose Estimation
  • Augmented Reality
  • Human-computer Interaction (HCI)

During my PhD study, I mainly focus on the vision-based human motion analysis, which aims to estimate the human pose and analysis the human motion from the RGB camera or RGB-D sensor (e.g. Kinect).

Currently, I am working on medical image analysis in the Lister Hill National Center for Biomedical Communications (LHNCBC), National Library of Medicine (NLM), National Institutes of Health (NIH), USA.

1. Real-time Human Pose Estimation from a Single Depth Sensor

human pose estimation

Hand Pose Estimation from a Single Depth Sensor

hand pose estimation

2. Articulated Gaussian Kernel Correlation for Similarity Measurement

2.1 Gaussian kernel correlation

Gaussian Kernel Correlation


2.2 Articulated structure

Articulation Pose Estimation


2.3 Generalized Gaussian kernel correlation

Gaussian Kernel Correlation
Generalized Gaussian Kernel Correlation

3. Manifold Learning for Human Motion Modeling

manifold learning
Human motion animation

Application for animation

Reference:

[1] Meng Ding, Guoliang Fan, "Articulated and Generalized Gaussian Kernel Correlation for Human Pose Estimation", in IEEE Transactions on Image Processing, vol.25, no.2, pp.776-789, Feb. 2016. [PDF available in IEEE Xplore]

[2] Meng Ding, Guoliang Fan, "Articulated Gaussian Kernel Correlation for Human Pose Estimation", in IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Boston, MA, USA, June 2015

[3] Meng Ding, Guoliang Fan,"Generalized Sum of Gaussians for Real-Time Human Pose Tracking from a Single Depth Sensor", in IEEE Winter Conference on Applications of Computer Vision (WACV), Big Island, Hawaii, USA, January 2015

[4] Meng Ding, Guoliang Fan, "Fast Human Pose Tracking with a Single Depth Sensor Using Sum of Gaussians Models", in International Symposium on Visual Computing (ISVC), Las Vegas, USA, 2014, published in Part I, LNCS 8887, pp. 599-608, 2014

Reference:

[1] Meng Ding, Guoliang Fan, "Articulated and Generalized Gaussian Kernel Correlation for Human Pose Estimation", in IEEE Transactions on Image Processing, vol.25, no.2, pp.776-789, Feb. 2016. [PDF available in IEEE Xplore]

Reference:

[1] Meng Ding, Guoliang Fan, "Multilayer Joint Gait-Pose Manifolds for Human Gait Motion Modeling", in IEEE Transactions on Cybernetics, vol.45, no. 11, pp.2413-2424. Nov. 2015 [pdf]

[2] Meng Ding, Guoliang Fan, “Multi-Layer Joint Gait-Pose Manifold for Human Motion Modeling”, in IEEE International Conference on Automatic Face & Gesture Recognition (FG), Shanghai, China, April 2013 (Spotlight). [pdf]

[3] Meng Ding, Guoliang Fan, Xin Zhang, et al., “Structure-guided Manifold Learning for Video-based Motion Estimation”, in IEEE International Conference on Image Processing (ICIP), Orlando, Florida, USA, October 2012. [pdf]

The techniques and tools I am working on includes:

  • Medical Image Segmentation
  • Gaussian Kernel Correlation, Articulated and Generalized Gaussian Kernel Correlation
  • Rigid, Non-rigid Registration
  • Unsupervised Learning, Clustering
  • Manifold Learning, Nonlinear Dimensional Reduction
  • SVM and Sparse Representation for Pattern Recognition
  • CNN, Caffe framework
  • Matlab, C / C++, Java, Python
  • OpenCV, OpenGL, OpenNI, Kinect SDK, CUDA
  • DICOM, DCMTK

Also, I did many research projects on Augmented Reality and Flight Simulator during my Master degree study.