Segmentation-free Word Spotting with Exemplar SVMs

Jon Almazán, Albert Gordo, Alicia Fornés, Ernest Valveny


In this paper we propose an unsupervised segmentation-free method for word spotting in document images. Documents are represented with a grid of HOG descriptors, and a sliding-window approach is used to locate the document regions that are most similar to the query. We use the Exemplar SVM framework to produce a better representation of the query in an unsupervised way. Then, we use a more discriminative representation based on Fisher Vector to rerank the best regions retrieved, and the most promising ones are used to expand the Exemplar SVM training set and improve the query representation. Finally, the document descriptors are precomputed and compressed with Product Quantization. This offers two advantages: first, a large number of documents can be kept in RAM memory at the same time. Second, the sliding window becomes significantly faster since distances between quantized HOG descriptors can be precomputed. Our results significantly outperform other segmentation-free methods in the literature, both in accuracy and in speed and memory usage.



[New] Extended version accepted in Pattern Recognition
Original BMVC2012 paper
BMVC2012 Poster


author = {J. Almaz\'an and A. Gordo and A. Forn\'es and E. Valveny},
title = {Segmentation-free Word Spotting with Exemplar SVMs},
booktitle = {Pattern Recognition}, 
year = {2014}}

author = {J. Almaz\'an and A. Gordo and A. Forn\'es and E. Valveny},
title = {Efficient Exemplar Word Spotting},
booktitle = {BMVC}, 
year = {2012}}