ICDAR2017 Competition on Historical Document Writer Identification (Historical-WI) (Historical-WI)

Research Tasks

Writer identification

2018-08-02 (v. 1)

Contact author

Stefan Fiel

TU Wien

fiel@cvl.tuwien.ac.at

+43 1 58801 18386


This work is licensed under a Creative Commons Attribution 3.0 Unported License.

Description

The first digits of the filename corresponds to the writer id.

The dataset is evaluated uses a leave-one-out strategy for evaluation. This means that a ranking according to the similarity of the handwriting has to be generated for a reference document (each document is taken once as reference document.

Protocol

The performance measurement used is the mean average precision (map), which is a standard evaluation metric for information retrieval (Wikipedia) because it not only measures the precision, but also the ranking itself. If the other 4 pages of an author ranked at the first 4 places for all documents, a map of 100% is achieved. Furthermore you can evaluate your method using the Soft and Hard Crituerion as described in the Competitions of 2011 and 2013. Also the percentage @ X is a usefull criterion which describes the percentage of correct documents in the first X documents of your ranking

Comments

No comments on this dataset yet.

Add your comment

In order to comment on a dataset you need to be logged on
Register Now!

Valoration

In order to rate this dataset you need to be logged on
Register Now!