9th Workshop "Theoretical Foundations of Computer Vision"
Evaluation and Validation of Computer Vision Algorithms
Schloß Dagstuhl/Wadern (near Saarbrücken/Germany)
March 16 - 20, 1998
This workshop addressed a subject which has been under active discussion in computer vision for several years. The evaluation and validation of algorithms is of basic importance for the configuration of computer vision applications.
"...the authors should clearly state their assumptions and justify their claims based on results that can be obtained under those assumptions. Vague justifications, such as subjective evaluations of images ... should not be allowed." (R.C. Jain, Th.O. Binford: Ignorance, myopia, and naivetÈ in computer vision systems. Computer Vision, Graphics, and Image Processing: Image Understanding 53 (1991), pp. 112 - 117. )In the ideal case certain "data sheets" should allow to qualify algorithmic solutions in a specific context, e.g. defined by image data, goal of image analysis, or software environment ("edge detection is not equal to edge detection"). There is a lack of methodological fundamentals in the field of performance analysis.
Participants presented ideas towards general methodology of evaluation and validation of algorithms or/and the application of these techniques to computer vision algorithms.
The proceedings of the workshop have been published: Klette, R., Stiehl, S., Viergever, M., Vincken, V. (eds.): Performance Evaluation of Computer Vision Algorithms. Kluwer Academic Publishers, Amsterdam 2000.
Related Web Sites
Algorithm Performance Contest, ICPR 2000
Workshops "Theoretical Foundations of Computer Vision":
Multi-Image Search, Filtering, Reasoning and Visualization (chairs:
Alfred Bruckstein, Thomas Huang, Reinhard Klette, SongDe Ma).
CITR: last update: 4 April 2000