Adaptation of an address reading system to local mail streams

Anja Brakensiek; Jörg Rottland; Frank Wallhoff; Gerhard Rigoll

doi:10.1109/ICDAR.2001.953911

Adaptation of an address reading system to local mail streams

Anja Brakensiek, Jörg Rottland, Frank Wallhoff, Gerhard Rigoll

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

5 Scopus citations

Abstract

In this paper a scheme for handwriting adaptation for post offices is described to improve recognition performance of German addresses. The recognition system is based on a tied-mixture Hidden Markov Model (HMM), whose parameters are updated using the expectation maximization (EM) technique, the maximum likelihood linear regression (MLLR) algorithm and a new discriminative adaptation technique, the scaled likelihood linear regression (SLLR). Contrary to the usual approach of adapting a writer-independent system to a specific writer, we propose here to adapt the system to the writer-independent data of a specific post office. The resulting system for each post office yields up to 16% lower word recognition errors.

Original language	English
Title of host publication	Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001
Publisher	IEEE Computer Society
Pages	872-876
Number of pages	5
ISBN (Electronic)	0769512631, 0769512631, 0769512631
DOIs	https://doi.org/10.1109/ICDAR.2001.953911
State	Published - 2001
Externally published	Yes
Event	6th International Conference on Document Analysis and Recognition, ICDAR 2001 - Seattle, United States Duration: 10 Sep 2001 → 13 Sep 2001

Publication series

Name	Proceedings of the International Conference on Document Analysis and Recognition, ICDAR
Volume	2001-January
ISSN (Print)	1520-5363

Conference

Conference	6th International Conference on Document Analysis and Recognition, ICDAR 2001
Country/Territory	United States
City	Seattle
Period	10/09/01 → 13/09/01

Access to Document

10.1109/ICDAR.2001.953911

Cite this

Brakensiek, A., Rottland, J., Wallhoff, F., & Rigoll, G. (2001). Adaptation of an address reading system to local mail streams. In Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001 (pp. 872-876). Article 953911 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR; Vol. 2001-January). IEEE Computer Society. https://doi.org/10.1109/ICDAR.2001.953911

@inproceedings{f6846c2fa18d43bba7156cb4ff0b28d6,

title = "Adaptation of an address reading system to local mail streams",

abstract = "In this paper a scheme for handwriting adaptation for post offices is described to improve recognition performance of German addresses. The recognition system is based on a tied-mixture Hidden Markov Model (HMM), whose parameters are updated using the expectation maximization (EM) technique, the maximum likelihood linear regression (MLLR) algorithm and a new discriminative adaptation technique, the scaled likelihood linear regression (SLLR). Contrary to the usual approach of adapting a writer-independent system to a specific writer, we propose here to adapt the system to the writer-independent data of a specific post office. The resulting system for each post office yields up to 16% lower word recognition errors.",

author = "Anja Brakensiek and J{\"o}rg Rottland and Frank Wallhoff and Gerhard Rigoll",

note = "Publisher Copyright: {\textcopyright} 2001 IEEE.; 6th International Conference on Document Analysis and Recognition, ICDAR 2001 ; Conference date: 10-09-2001 Through 13-09-2001",

year = "2001",

doi = "10.1109/ICDAR.2001.953911",

language = "English",

series = "Proceedings of the International Conference on Document Analysis and Recognition, ICDAR",

publisher = "IEEE Computer Society",

pages = "872--876",

booktitle = "Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001",

}

Brakensiek, A, Rottland, J, Wallhoff, F & Rigoll, G 2001, Adaptation of an address reading system to local mail streams. in Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001., 953911, Proceedings of the International Conference on Document Analysis and Recognition, ICDAR, vol. 2001-January, IEEE Computer Society, pp. 872-876, 6th International Conference on Document Analysis and Recognition, ICDAR 2001, Seattle, United States, 10/09/01. https://doi.org/10.1109/ICDAR.2001.953911

Adaptation of an address reading system to local mail streams. / Brakensiek, Anja; Rottland, Jörg; Wallhoff, Frank et al.
Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001. IEEE Computer Society, 2001. p. 872-876 953911 (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR; Vol. 2001-January).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Adaptation of an address reading system to local mail streams

AU - Brakensiek, Anja

AU - Rottland, Jörg

AU - Wallhoff, Frank

AU - Rigoll, Gerhard

PY - 2001

Y1 - 2001

N2 - In this paper a scheme for handwriting adaptation for post offices is described to improve recognition performance of German addresses. The recognition system is based on a tied-mixture Hidden Markov Model (HMM), whose parameters are updated using the expectation maximization (EM) technique, the maximum likelihood linear regression (MLLR) algorithm and a new discriminative adaptation technique, the scaled likelihood linear regression (SLLR). Contrary to the usual approach of adapting a writer-independent system to a specific writer, we propose here to adapt the system to the writer-independent data of a specific post office. The resulting system for each post office yields up to 16% lower word recognition errors.

AB - In this paper a scheme for handwriting adaptation for post offices is described to improve recognition performance of German addresses. The recognition system is based on a tied-mixture Hidden Markov Model (HMM), whose parameters are updated using the expectation maximization (EM) technique, the maximum likelihood linear regression (MLLR) algorithm and a new discriminative adaptation technique, the scaled likelihood linear regression (SLLR). Contrary to the usual approach of adapting a writer-independent system to a specific writer, we propose here to adapt the system to the writer-independent data of a specific post office. The resulting system for each post office yields up to 16% lower word recognition errors.

UR - http://www.scopus.com/inward/record.url?scp=35048825729&partnerID=8YFLogxK

U2 - 10.1109/ICDAR.2001.953911

DO - 10.1109/ICDAR.2001.953911

M3 - Conference contribution

AN - SCOPUS:35048825729

T3 - Proceedings of the International Conference on Document Analysis and Recognition, ICDAR

SP - 872

EP - 876

BT - Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001

PB - IEEE Computer Society

T2 - 6th International Conference on Document Analysis and Recognition, ICDAR 2001

Y2 - 10 September 2001 through 13 September 2001

ER -

Brakensiek A, Rottland J, Wallhoff F, Rigoll G. Adaptation of an address reading system to local mail streams. In Proceedings - 6th International Conference on Document Analysis and Recognition, ICDAR 2001. IEEE Computer Society. 2001. p. 872-876. 953911. (Proceedings of the International Conference on Document Analysis and Recognition, ICDAR). doi: 10.1109/ICDAR.2001.953911

Adaptation of an address reading system to local mail streams

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this