Skip to Main content Skip to Navigation
Conference papers

On the Use of Attention Mechanism in a Seq2Seq based Approach for Off-line Handwritten Digit String Recognition

Abstract : In this work, we investigate the use of the attention mechanism in deep learning for a better reading of handwritten digit strings in digitized images. The proposed recognition system built upon a CNN (Convolutional Neural Network) and two RNNs (Recurrent Neural Networks), acting as Encoder and Decoder and using the attention mechanism. We used a 1D mechanism for attention location with a "soft" alignment attention which has the peculiarity of having an easily calculable gradient and thus to integrate well with the network. Experimental results on data from ORAND-CAR A, ORAND-CAR B and CVL HDS databases compare favorably to other published methods.
Document type :
Conference papers
Complete list of metadata

Cited literature [16 references]  Display  Hide  Download

https://hal.inria.fr/hal-02460896
Contributor : Abdel Belaid <>
Submitted on : Thursday, January 30, 2020 - 12:36:39 PM
Last modification on : Thursday, June 17, 2021 - 3:02:33 AM

File

Thiebault-ICDAR.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02460896, version 1

Collections

Citation

T Lupinski, A. Belaid, A Echi. On the Use of Attention Mechanism in a Seq2Seq based Approach for Off-line Handwritten Digit String Recognition. ICDAR, Sep 2019, Sydney, Australia. ⟨hal-02460896⟩

Share

Metrics

Record views

81

Files downloads

693