Word Spotting and Recognition with Embedded Attributes

Welcome to the Word Representation with Attributes library, a software for the retrieval and recognition of word images.

This code is written in Matlab and is the basis of the following project:

Jon Almazán, Albert Gordo, Alicia Fornés, Ernest Valveny. Word Spotting and Recognition with Embedded Attributes. ICCV2013. Project page

Abstract

We deal with the problems of word spotting and word recognition on images. In word spotting, the goal is to find all instances of a query word in a dataset of images. In recognition, the goal is to recognize the content of the word image, usually aided by a dictionary or lexicon. We propose a formulation for word representation and matching based on embedded attributes that jointly addresses these two problems. Contrary to most other existing methods, our representation has a fixed length, is low dimensional, and is very fast to compute and, especially, to compare.

We propose to use character attributes to learn a semantic representation of the word images and then perform a calibration of the scores with CCA that puts images and text strings in a common subspace. After that, spotting and recognition become simple nearest neighbor problems in a very low dimensional space. We test our approach on four public datasets of both document and natural images showing results comparable or better than the state-of-the-art on spotting and recognition tasks.

This word spotting library uses great open-source software:

• VLFeat

---

MATLAB Quick Start Guide

To get started, you need to install MATLAB and download the code from GitHub. This code has been tested on Mac and Linux and some pre-compiled Mex files are included.

Download source code

$ cd ~/your_projects/
$ git clone git://github.com/almazan/watts.git

Download and uncompress the IIIT5K dataset

$ cd watts/datasets
$ wget http://cvit.iiit.ac.in/projects/SceneTextUnderstanding/IIIT5K-Word_V3.0.tar.gz
$ tar -xvzf IIIT5K-Word_V3.0.tar.gz

Download precomputed models for IIIT5K (optional)

$ cd watts/data
$ wget http://www.cvc.uab.es/~almazan/data/IIIT5K_models.tar.gz
$ tar -xvzf IIIT5K_models.tar.gz

Run the program with the default parameters

>> main

Note: The default parameters as well as the dataset selection and paths can be modified in the prepare_opts.m script.

---

Authors

The code has been developed by @almazan and @agordo.

License

This code has been released under the MIT license.