Publication: Sparse Radial Sampling LBP for Writer Identification


Authors: Anguelos Nicolaou, Andrew D. Bagdanov, Marcus Liwicki, and Dimosthenis Karatzas


Published in: ICDAR 2015 (oral presentation)


Download pdf


Abstract:
In this paper we present the use of Sparse Radial Sampling Local Binary Patterns, a variant of Local Binary Patterns (LBP) for text-as-texture classification. By adapting and extending the standard LBP operator to the particularities of text we get a generic text-as-texture classification scheme and apply it to writer identification. In experiments on CVL and ICDAR 2013 datasets, the proposed feature-set and a simple end-to-end pipeline demonstrate State-Of-the-Art (SOA) performance. Among the SOA, the proposed method is the only one that is based on dense extraction of a single local feature descriptor. This makes it fast and applicable at the earliest stages in a DIA pipeline without the need for segmentation, binarization, or extraction of multiple features.


Aditional Resources:

slides presentation.pdf

Experimental data and reproducabillity:

In order to allow for the reproducabillity of most experiments described in the article, SRS-LBP histograms on the most recent datasets are provided in csv format.

Sample usage:

#Reproducing leave-one-out measurement on ICDAR 2013 dataset
python srs_wi_pipeline.py -r 0 1 2 3 4 5 6 7 8 9 10 11 -s bilinear -T otsu -p pca -m l1out -q icdar2013_bilinear_r1to20_otsu.csv 
#Reproducing metric measurement on ICDAR 2013 dataset, learning pca from the ICFHR 2012 Dataset
python srs_wi_pipeline.py -r 0 1 2 3 4 5 6 7 8 9 10 11 -s bilinear -T otsu -p pca -m l1out -q icdar2013_bilinear_r1to20_otsu.csv -s  icfhr2012_bilinear_r1to20_otsu.csv 

bibtex entry :
 @article{nicolaou2015sparse,
title={Sparse Radial Sampling LBP for Writer Identification},
author={Nicolaou, Anguelos and Bagdanov, Andrew D and Liwicki, Marcus and Karatzas, Dimosthenis},
journal={arXiv preprint arXiv:1504.06133},
year={2015}
}


Creative Commons License
Any work in this page other than source code or program binaries is licensed under a Creative Commons Attribution 4.0 International License. When applicable atribution should be in the form of a citation.