Year: 2006
Type: article
Abstract: Classifiers for object categorization and identification, such as face detectors and face recognizers, are often trained separately and operated in a feed-forward fashion. Selecting a small number of features for these tasks is important to prevent over-fitting and reduce computation. However, when a system has such related or sequential tasks, training and selecting features for these tasks independently may not be optimal. We propose a framework for choosing features to be shared between categorization and identification tasks. The result is a system that achieves better performance with a given number of features. We demonstrate with experiments using text and car detection as categorization tasks, and character and vehicle type recognition as identification tasks. 1.
Institution: University of Massachusetts Amherst
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2110648765