Hough-based Object Detection with Grouped Features

2014

Conference Paper

ps


Hough-based voting approaches have been successfully applied to object detection. While these methods can be efficiently implemented by random forests, they estimate the probability for an object hypothesis for each feature independently. In this work, we address this problem by grouping features in a local neighborhood to obtain a better estimate of the probability. To this end, we propose oblique classification-regression forests that combine features of different trees. We further investigate the benefit of combining independent and grouped features and evaluate the approach on RGB and RGB-D datasets.

Author(s): Abhilash Srikantha and Juergen Gall
Book Title: International Conference on Image Processing
Pages: 1653--1657
Year: 2014
Month: October

Department(s): Perceiving Systems
Research Project(s): Object Detection
Bibtex Type: Conference Paper (conference)
Paper Type: Conference

Address: Paris, France
DOI: http://dx.doi.org/10.1109/ICIP.2014.7025331
Event Name: IEEE International Conference on Image Processing
Event Place: Paris, France
Attachments: pdf
poster

BibTex

@conference{Srikantha:ICIP:2014,
  title = {Hough-based Object Detection with Grouped Features},
  author = {Srikantha, Abhilash and Gall, Juergen},
  booktitle = {International Conference on Image Processing},
  pages = {1653--1657},
  address = {Paris, France},
  month = oct,
  year = {2014},
  month_numeric = {10}
}