Deep salience map guided arbitrary direction scene text recognition

Xinhao Liu; Takahito Kawanishi; Xiaomeng Wu; Kaoru Hiramatsu; Kunio Kashino

ICASSP 2017accepted0 citations

Deep salience map guided arbitrary direction scene text recognition

Xinhao Liu, Takahito Kawanishi, Xiaomeng Wu, Kaoru Hiramatsu, Kunio Kashino

Abstract

Irregular scene text such as curved, rotated or perspective texts commonly appear in natural scene images due to different camera view points, special design purposes etc. In this work, we propose a text salience map guided model to recognize these arbitrary direction scene texts. We train a deep Fully Convolutional Network (FCN) to calculate the precise salience map for texts. Then we estimate the positions and rotations of the text and utilize this information to guide the generation of CNN sequence features. Finally the sequence is recognized with a Recurrent Neural Network (RNN) model. Experiments on various public datasets show that the proposed approach is robust to different distortions and performs superior or comparable to the state-of-the-art techniques.

BibTeX

@inproceedings{icassp2017_deepsaliencemapg,
  title = {Deep salience map guided arbitrary direction scene text recognition},
  author = {Xinhao Liu and Takahito Kawanishi and Xiaomeng Wu and Kaoru Hiramatsu and Kunio Kashino},
  booktitle = {ICASSP 2017},
  year = {2017}
}