Zhiding Yu

 

 

Ph.D. Candidate

Dept. of Electrical and Computer Engineering, Carnegie Mellon University
B200 Wing, Hamerschlag Hall, Carnegie Mellon University

5000 Forbes Ave, Pittsburgh, PA 15213

Advisor: Prof. Vijayakumar Bhagavatula
 

Email: yzhiding AT andrew.cmu.edu

I'm looking for 2013 Summer Intern related to Computer Vision Research. The Curriculum Vitae is available here: [PDF]

Biography

I am currently a Ph.D. candidate at the Department of Electrical and Computer Engineering, Carnegie Mellon University. Before that I obtained the M.Phil. degree from the Department of Electrical and Computer Engineering, The Hong Kong University of Science and Technology.

My research interests mainly focus on Graph-Embedded Mode Seeking, Image Segmentation, Object Recognition and Scene Understanding.

I am the awardee of  the 2009-2010/2011-2012 HKTIIT Post-Graduate Excellence Scholarships and Carnegie Institute of Technology Dean's Tuition Fellowship.

Educational Background

Selected Publications

Thesis

Journal Papers

Conference Papers

Selected Projects

Multi-Class Object Semantic Segmentation

 

 I am particularly interested in joint object recognition and segmentation. I am currently investigating some interesting ideas on this topic. Here are some premative results.

Automatic Clear Path Detection for Autonomous Driving

The clear path detection method for car reversing intended to be developed in this project is basically three-folded: It consists 1. building clear path models using both offline learning and online learning, 2. false positive rejection by pedestrian and vehicle detectors, and 3. final inference of clear path labels.

Differs from traditional clear path detection methods, our method takes the advantage of available priors in a car reversing scenario, particularly with online training which can improve the classification accuracy. The algorithm will operate on the superpixel framework for speed up and better feature extraction purposes. Each superpixel will be assigned a label to indicate whether it is clear path or not.

A premitive result video is available here.

Tree Embedded Mode Seeking for Manifold Structured Data Clustering and Image Segmentation

We present a novel framework for tree-structure embedded density estimation and its fast approximation for mode seeking. Given any undirected, connected and weighted graph, the density function is defined as a joint representation of the feature space and the distance domain on the graph’s spanning tree. Since the distance domain of a tree is a constrained one, mode seeking can not be directly achieved by traditional mean shift in both domain. we address this problem by introducing node shifting with force competition and its fast approximation. The new formulation of this problem can lead to many advantages and new characteristics in its application.

This work appears in CVPR 2011.

Automatic Object Segmentation from Large Scale 3D Urban Point Clouds


 

We present a system that can automatically segment objects in large scale 3D point clouds obtained from urban ranging images. The system consists of three steps: The first one involves a ground detection process that can detect relatively complex terrain and separate it from other objects. The second step superpixelizes the remaining objects to speed up the segmentation process. In the final step, a manifold embedded mode seeking method is adopted to segment the point clouds. Even though the segmentation of urban objects is a challenging problem in terms of accuracy and problem scale, our system can efficiently generate very good segmentation results. The proposed manifold learning effectively improves the segmentation performance due to the fact that continuous artificial objects often have manifold-like structures.

This work appears in ACM-MM 2011.

ELEC 547 Convex Optimization Course Project: Mode Seeking with Convex Shift and Interactive Learning

  
 

  

We give a brief introduction of nonparametric density estimation and the family of mode seeking methods under this framework. Regarding the limit of current methods, we propose several improvements. The major contribution is basically two fold:

1. We show that for mean shift with a linear kernel, each kernel shift can actually be formulated in a convex form. By doing this we further generalize it for several convex metrics - e.g., KL divergence and Jeffrey divergence - where it is difficult or impossible to clearly define a ”mean” for mode seeking.

2. We also introduce interactive learning to image segmentation which is a specific application of mode seeking algorithms. By adopting interactive segmentation through convex learning, better segmentation results can be achieved over segmentation with traditional mode seeking algorithms.

The project report is available here.

An improved version of this work later appeared in CVPR 2012.

Honors and Awards

More Information About Me

Friends & Collaborators

 

This page has been visited

 图片计数器

times