Pilot
European
Image
Processing
Archive
The PCCV Project:
Benchmarking Vision Systems
Overview
Tutorials
Methodology
Case studies
Test datasets
Our image file format
HATE test harness
Information
General links
Conferences
Mailing lists
Research groups
Societies
Techniques (CVonline)
Software
Image databases
Other stuff
Linux on ThinkPad
Introduction
The reasons for evaluating the performances of algorithms have been outlined on the PEIPA home page and examined in detail in the benchmarking pages and associated tutorials. The following discussion assumes that you are considering carrying out some evaluation or comparison yourself and are considering using HATE to make your life easier.
HATE stands for Harness for Algorithm Testing and Evaluation. As its name suggests, it is a test harness: a tool that is used to automate the process of applying tests to a piece of software and accumulating the results. It can be applied to pretty well any type of testing, but was designed and implemented with characterizing the performance of vision algorithms very much in mind.
HATE is currently undergoing a major programme of extensions, and the way the new version is used will differ somewhat from earlier versions. For that reason, the material describing it has currently been removed, in anticipation of material concerning the new version being made available in February 2004. But if you can't wait to get your hand on HATE, please contact the maintainers!