Discriminative Learning for Deformable Shape Segmentation: A Comparative Study

  • Authors:

  • Jingdan Zhang;Shaohua Kevin Zhou;Dorin Comaniciu;Leonard Mcmillan

  • Affiliations:

  • Integrated Data Systems Department, Siemens Corporate Research, Princeton, USA NJ 08540 and Department of Computer Science, UNC Chapel Hill, Chapel Hill, USA NC 27599;Integrated Data Systems Department, Siemens Corporate Research, Princeton, USA NJ 08540;Integrated Data Systems Department, Siemens Corporate Research, Princeton, USA NJ 08540;Department of Computer Science, UNC Chapel Hill, Chapel Hill, USA NC 27599

  • Venue:
  • ECCV '08 Proceedings of the 10th European Conference on Computer Vision: Part I
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a comparative study on how to use discriminative learning methods such as classification, regression, and ranking to address deformable shape segmentation. Traditional generative models and energy minimization methods suffer from local minima. By casting the segmentation into a discriminative framework, the target fitting function can be steered to possess a desired shape for ease of optimization yet better characterize the relationship between shape and appearance. To address the high-dimensional learning challenge present in the learning framework, we use a multi-level approach to learning discriminative models. Our experimental results on left ventricle segmentation from ultrasound images and facial feature point localization demonstrate that the discriminative models outperform generative models and energy minimization methods by a large margin.