1. Overview of Clustering with Non-parametric NPclassify

Overview of clustering with NPclassify (Non Parametric Classify) and how it works.

2. Basic Validation

NPclassify is compared with K-means for validation purposes.

3. Training Images

This is the set of images used to train for feature based clustering.

4. Validation Images

This is the set of images used to validate the model.

5. Results for Segmentation

These are the resulting clustered features from the images in (2).

6. Results for Classification

These are features clustered across 10 images using the same settings as (3) but combining features across images.

7. Download the Source Code

The soource code is GPL and is part of the iLab Neuromorphic Vision Toolkit.

8. Publication describing NPclassify clustering

Biologically inspired feature based categorization of objects, T. N. Mundhenk, V. Navalpakkam, H. Makaliwe, S. Vasudevan, L. Itti, Proc. SPIE Human Vision and Electronic Imaging IX,  San Jose , California, January 2004 [5292-29]

Other publications relating to NPclassify:

Teaching the computer subjective notions of feature connectedness in a visual scene for real time vision, T. N. Mundhenk, C. Landauer, K. Bellman, M. A. Arbib, L. Itti, Proc. SPIE Conference on Intelligent Robots and Computer Vision XXII, Philadelphia, PA, October 2004

Distributed biologically based real time tracking in the absence of prior target information, T. N. Mundhenk, J. Everist, C. Landauer, L. Itti, K. Bellman, Proc SPIE Conference on Intelligent Robots and Computer Vision XXIII, Boston, MA, October 2005

| iLab Home | Nerd-Cam Home | CINNIC Home |

| Mundhenk.com | Beobot Home |

Copyright 2004, T. Nathan Mundhenk, iLab, University of Southern California