A statistical approach for phrase location and recognition within a text line: an application to street name recognition

Yacoubi, Mounîm El; Gilloux, M.; Bertille, J.-M.

doi:10.1109/34.982898

Cited by 31 publications

(16 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is based on Hidden Markov Models (HMMs). HMMs are state-of-the-art for modeling handwritten text [51] and have been widely used for keyword spotting [3], [28], [31], [33], [47], [52], [53]. In [30], trained character models are used to spot arbitrary keywords in complete text line images using an efficient lexicon-free approach.…”

Section: B Hmm Reference Systemmentioning

confidence: 99%

A Novel Word Spotting Method Based on Recurrent Neural Networks

Frinken

Fischer

Manmatha

et al. 2012

IEEE Trans. Pattern Anal. Mach. Intell.

242

191

View full text Add to dashboard Cite

Abstract-Keyword spotting refers to the process of retrieving all instances of a given keyword from a document. In the present paper, a novel keyword spotting method for handwritten documents is described. It is derived from a neural network based system for unconstrained handwriting recognition. As such it performs template-free spotting, i.e. it is not necessary for a keyword to appear in the training set. The keyword spotting is done using a modification of the CTC Token Passing algorithm in conjunction with a recurrent neural network. We demonstrate that the proposed systems outperforms not only a classical dynamic time warping based approach but also a modern keyword spotting system, based on hidden Markov models. Furthermore, we analyze the performance of the underlying neural networks when using them in a recognition task followed by keyword spotting on the produced transcription. We point out the advantages of keyword spotting when compared to classic text line recognition.

show abstract

Section: B Hmm Reference Systemmentioning

confidence: 99%

A Novel Word Spotting Method Based on Recurrent Neural Networks

Frinken

Fischer

Manmatha

et al. 2012

IEEE Trans. Pattern Anal. Mach. Intell.

242

191

View full text Add to dashboard Cite

show abstract

“…Hidden Markov Models are double stochastic models that have been shown to be suitable for modeling, analyzing and recognizing sequential data such speech [15], handwriting [8], bioinformatics [19] and gestures. HMMs are a natural choice to model the gait data.…”

Section: Hmm Modeling and Recognitionmentioning

confidence: 99%

HMM-based gait modeling and recognition under different walking scenarios

Yacoubi

Shaiek

Dorizzi

2011

2011 International Conference on Multimedia Computing and Systems

View full text Add to dashboard Cite

International audienceThis paper addresses gait recognition, the problem of identifying people by the way of their walk. The proposed system consists of a model-free approach which extracts features directly from the human silhouette. The dynamics of the gait are modeled using Hidden Markov Models. Experiments have been carried out on the CASIA dataset C consisting of 153 people under four walking scenarios: normal walking, slow walking, fast walking and walking while carrying a bag. The results obtained are promising and compare favorably with existing approache

show abstract

“…In other related recognition domains where the vocabulary of the input is potentially very large, ergodic topologies have also been proposed. We can refer to [7] where an ergodic HMM system is presented to recognize handwritten street names, to [16] for automatic language identification and to [12] for speaker verification.…”

Section: Recognition With Ergodic Topologymentioning

confidence: 99%

A language-independent, open-vocabulary system based on HMMs for recognition of ultra low resolution words

Einsele

Ingold

Hennebert

2008

Proceedings of the 2008 ACM Symposium on Applied Computing

View full text Add to dashboard Cite

In this paper, we introduce and evaluate a system capable of recognizing ultra low resolution words extracted from images such as those frequently embedded on web pages. The design of the system has been driven by the following constraints. First, the system has to recognize small font sizes where antialiasing and resampling procedures have been applied. Such procedures add noise on the patterns and complicate any a priori segmentation of the characters. Second, the system has to be able to recognize any words in an open vocabulary setting, potentially mixing different languages. Finally, the training procedure must be automatic, i.e. without requesting to extract, segment and label manually a large set of data. These constraints led us to an architecture based on ergodic HMMs where states are associated to the characters. We also introduce several improvements of the performance increasing the order of the emission probability estimators and including minimum and maximum duration constraints on the character models. The proposed system is evaluated on different font sizes and families, showing good robustness for sizes down to 6 points.

show abstract

A statistical approach for phrase location and recognition within a text line: an application to street name recognition

Cited by 31 publications

References 26 publications

A Novel Word Spotting Method Based on Recurrent Neural Networks

A Novel Word Spotting Method Based on Recurrent Neural Networks

HMM-based gait modeling and recognition under different walking scenarios

A language-independent, open-vocabulary system based on HMMs for recognition of ultra low resolution words

Contact Info

Product

Resources

About