Writer Identification Using Super Paramagnetic Clustering and Spatio Temporal Neural Network

Authors:
Seyyed Ataollah Taghavi Sangdehi;Karim Faez
Affiliations:
Qazvin Azad University of Iran, Amirkabir University, Tehran, Iran;Qazvin Azad University of Iran, Amirkabir University, Tehran, Iran
Venue:
CIARP '09 Proceedings of the 14th Iberoamerican Conference on Pattern Recognition: Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications
Year:
2009

Citing 2
Cited 0

Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering

Neural Computation
On-line signature recognition based on VQ-DTW

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper discusses use of Super Paramagnetic Clustering (SPC) and Spatio Temporal Artificial Neuron in on-line writer identification, on Farsi handwriting. In online cases, speed and automation are advantages of one method on others, therefore we used unsupervised and relatively quick clustering method, which in comparison with conventional approaches, give us better result. Moreover, regardless of various parameters that available from acquisition systems, we only consider to displacement of pen tip at determined direction that lead to quick system due to its quick preprocessing and clustering. Also we use a threshold that remove displacement between disconnected point of a word that lead to a better classification result on on-line Farsi writers.