Ground-Truth Production in the Transcriptorium Project

Gatos, Basilis; Louloudis, Georgios; Causer, Tim; Grint, Kris; Romero, Verónica; Lafuente, José Vicente; Toselli, Alejandro H.; Vidal, Enrique

doi:10.1109/das.2014.23

Cited by 36 publications

(30 citation statements)

References 8 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These pages entailed several line detection and transcription difficulties and the corresponding ground truth (GT) was produced semi-automatically and manually reviewed [12] (see examples of extracted lines in Fig. 2).…”

Section: Dataset Descriptionmentioning

confidence: 99%

ICFHR2016 Competition on Handwritten Text Recognition on the READ Dataset

Lafuente

Romero

Toselli

et al. 2016

2016 15th International Conference on Frontiers in Handwriting Recognition (ICFHR)

Self Cite

View full text Add to dashboard Cite

IEEESánchez Peiró, JA.; Romero Gómez, V.; Toselli ., AH.; Vidal, E. (2016) The handwritten images for this competition were drawn from the German document Ratsprotokolle collection composed of minutes of the council meetings held from 1470 to 1805, used in the READ project. The selected dataset is written by several hands and entails significant variabilities and difficulties. The five participants achieved good results with transcriptions word error rates ranging from 21% to 47% and character error rates rating from 5% to 19%.

show abstract

Section: Dataset Descriptionmentioning

confidence: 99%

ICFHR2016 Competition on Handwritten Text Recognition on the READ Dataset

Lafuente

Romero

Toselli

et al. 2016

2016 15th International Conference on Frontiers in Handwriting Recognition (ICFHR)

Self Cite

View full text Add to dashboard Cite

show abstract

“…It is also difficult from the HTR point of view because it is written by several hands, it has crossed out words, hyphenated words, etc. Preliminary HTR results on a small set of 53 pages the Bentham collection were reported in [8] by using the HTR techniques mentioned in Section 2. The Word Error Rate (WER) obtained in that set was about 34% (see [8] for additional details).…”

Section: Current Collections Selected In Transcriptoriummentioning

confidence: 99%

“…Although these technologies are already providing useful results in some cases, much remains to be developed, especially for historical documents, which suffer from typical degradations [10,13,8].…”

Section: Introductionmentioning

confidence: 99%

Handwritten text recognition for historical documents in the transcriptorium project

Lafuente

Bosch

Romero

et al. 2014

Proceedings of the First International Conference on Digital Access to Textual Cultural Heritage

View full text Add to dashboard Cite

Transcription of historical handwritten documents is a crucial problem for making easier the access to these documents to the general public. Currently, huge amount of historical handwritten documents are being made available by on-line portals worldwide. It is not realistic to obtain the transcription of these documents manually, and therefore automatic techniques has to be used. tranScriptorium is a project that aims at researching on modern Handwritten Text Recognition (HTR) technology for transcribing historical handwritten documents. The HTR technology used in tranScriptorium is based on models that are learnt automatically from examples. This HTR technology has been used on a Dutch collection from 15th century selected for the tranScriptorium project. This paper provides preliminary HTR results on this Dutch collection that are very encouraging, taken into account that minimal resources have been deployed to develop the transcription system.

show abstract

“…The Bentham Dataset [10] consists of a series of documents from the Bentham collection. It has been prepared in the tranScriptorium project 1 .…”

Section: A Datasetmentioning

confidence: 99%

Visual Analysis System for Features and Distances Qualitative Assessment: Application to Word Image Matching

Rayar

Mondal

Barrat

et al. 2016

2016 12th IAPR Workshop on Document Analysis Systems (DAS)

View full text Add to dashboard Cite

Abstract-In this paper, a visual analysis system to qualitatively assess the features and distance functions that are used for calculating dissimilarity between two word images is presented. Computation of dissimilarity between two images is the prerequisite for image matching, indexing and retrieval problems. First, the features are extracted from the word images and a distance between each image to others is computed and represented in a matrix form. Then, based on this distance matrix, a proximity graph is built to structure the set of word images and highlight their topology. The proposed visual analysis system is a web based platform that allows visualisation and interactions on the obtained graph. This interactive visualisation tool inherently helps users to quickly analyse and understand the relevance and robustness of selected features and corresponding distance function in a unsupervised way, i.e. without any ground truth. Experiments are performed on a handwritten dataset of segmented words. Three types of features and four distance functions are considered to describe and compare the word images. Theses material are leveraged to evaluate the relevance of the built graph, and the usefulness of the platform.

show abstract

Ground-Truth Production in the Transcriptorium Project

Cited by 36 publications

References 8 publications

ICFHR2016 Competition on Handwritten Text Recognition on the READ Dataset

ICFHR2016 Competition on Handwritten Text Recognition on the READ Dataset

Handwritten text recognition for historical documents in the transcriptorium project

Visual Analysis System for Features and Distances Qualitative Assessment: Application to Word Image Matching

Contact Info

Product

Resources

About