ICDAR2019 COMPETITION ON IMAGE RETRIEVAL FOR HISTORICAL HANDWRITTEN DOCUMENTS
ICDAR-2019-HDRC-IR
#
Team Name
Best mAP up to now
Last submission mAP
0
Srslbp-l1out
82.97%
82.97 %
1
Srslbp-17
73.93%
73.93 %
Description:
This is a baseline method base on Sparse Radial Sampling Local Binary Patterns SRS-LBP.
For more details, refer, to Sparse Radial Sampling LBP for Writer Identification download.
In this variant, defferent PCA components were extracted from the retrieval database for every query and applied to both query and retrieval database.
Submissions:
mAP:
82.97
Precision:
44.28
Recall:
88.56
F-score:
59.04
Description:
This is a baseline method base on Sparse Radial Sampling Local Binary Patterns SRS-LBP.
For more details, refer, to Sparse Radial Sampling LBP for Writer Identification download.
In this variant The PCA components were extracted from the colored test set of the ICDAR2017 Competition on Historical Document Writer Identification.
