Synthetic On-line Handwriting Generation by Distortions and Analogy

Harold Mouchère 1 Sabri Bayoudh 2 Eric Anquetil 1 Laurent Miclet 2
1 IMADOC - Interprétation et Reconnaissance d’Images et de Documents
UR1 - Université de Rennes 1, INSA Rennes - Institut National des Sciences Appliquées - Rennes, CNRS - Centre National de la Recherche Scientifique : UMR6074
2 CORDIAL - Human-machine spoken dialogue
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, INRIA Rennes, ENSSAT - École Nationale Supérieure des Sciences Appliquées et de Technologie
Abstract : One of the difficulties to improve on the fly writer-dependent handwriting recognition systems is the lack of data available at the beginning of the adapting phase. In this paper we explore three possible strategies to generate synthetic handwriting characters from few samples of a writer. We explore in this paper both classical image distortions and two original ways to generate on-line handwritten characters: distortions based on specificities of the on-line handwriting and a generation based on analogical proportion. The experimentations show that these three approaches generate different distortions which are complementary. Indeed the combination of them allows to achieve using only 4 original characters for the learning phase a mean of 91.3% of recognition rate for 12 writers.
Complete list of metadatas

Cited literature [7 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/inria-00300700
Contributor : Harold Mouchère <>
Submitted on : Wednesday, January 16, 2019 - 5:44:08 PM
Last modification on : Thursday, February 7, 2019 - 3:29:53 PM

File

IGS07_final_version.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00300700, version 2

Citation

Harold Mouchère, Sabri Bayoudh, Eric Anquetil, Laurent Miclet. Synthetic On-line Handwriting Generation by Distortions and Analogy. in 13th Conference of the International Graphonomics Society (IGS2007), Nov 2007, Melbourne, Australia. pp.10-13. ⟨inria-00300700v2⟩

Share

Metrics

Record views

422

Files downloads

24