Word image matching using dynamic time warping

By Toni M. Rath, Raghavan Manmatha. 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings., 2003

675 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/e20e64661424bee772c7f002d45c1ac79578fb03

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Word Spotting in the Wild (2010, background): “The works that are most similar to ours are that of [13] and [5].”
Highly Comparative Feature-Based Time-Series Classification (2014, neutral): “2.3.”
Weighted dynamic time warping for time series classification (2011, methodology): “originally developed for speech recognition, but several researchers have evaluated its application in other domains and have developed several variants such as derivative DTW (DDTW) [11,21,22].”
Effective and General Evaluation for Instruction Conditioned Navigation using Dynamic Time Warping (2019, background): “DTW is a similarity function for time series that has long been known and used in speech processing [12, 13, 14], robotics [15, 16], data mining [17, 18], handwriting recognition [19], gesture recognition [20, 21] and many other areas [22, 23, 24].”
Exact indexing of dynamic time warping (2002, methodology): “Rath and Manmatha have applied DTW to the problem of indexing repositories of handwritten historical documents (Rath and Manmatha 2002) (although handwriting is two-dimensional, it can trivially be rerepresented as a one-dimensional time series).”
Computing and Visualizing Dynamic Time Warping Alignments in R: The dtw Package (2009, neutral): “…analysis (Huang and Kinsner 2002; Syeda-Mahmood et al. 2007; Tuzcu and Nas 2005), clustering of gene expression proﬁles (Aach and Church 2001; Hermans and Tsiporkova 2007), biometrics (Faundez-Zanuy 2007; Rath and Manmatha 2003), process monitoring (Gollmer and Posten 1996), just to name a few.”
Label embedding for text recognition (2013, neutral): “As an alternative to this bottom-up approach, a number of previous works have shown good results for small-vocabulary tasks by performing matching at full-image level, employing global features [4] or feature sequences [21, 22] and using a distance-based classification (feasibility for not so small ”
Word spotting for historical documents (2007, methodology): “If thresholding applied to the features extracted from the query and a candidate image determines that the images are dissimilar, we do not assign a similarity score to the candidate image (see [26] for details of this process).”
Lexicon-free handwritten word spotting using character HMMs (2012, methodology): “On the one hand, templatebased methods [9, 10, 11, 12, 13, 14, 15, 16, 17, 18] match a query word image with labeled keyword template images.”
Tight Hardness Results for LCS and Other Sequence Similarity Measures (2015, background): “Because of its generality, DTWD has been successfully applied in a large variety of domains: automatic speech recognition [43], music information retrieval [39] , bioinformatics [1], medicine [16], identifying songs from humming [50], indexing of historical handwriting archives [45], databases [35],”
Advances in Geocomputation: Geocomputation 2015--The 13th International Conference (2015, neutral): “In: Anselin L, Florax R, Sergio J (eds) Advances in spatial econometrics: methodology, tools and applications. Springer, Berlin, pp 121–142 Church RL, Marston JR (2003) Measuring accessibility for people with a disability. Geogr Anal 35 (1):83–96 Clayton DG (1996) Generalised linear mixed models.”
kml and kml3d: R Packages to Cluster Longitudinal Data (2015, background): “If speciﬁed, distance should be a function that takes two trajectories and returns a number, letting the user compute a non-classic distance (such as the adaptive dissimilarity index, Hennig and Hausdorf 2006; dynamic time warping, Rath and Manmatha 2003; Fr´echet distance, Fr´echet 1905; . . . ).”
Voice Recognition Algorithms using Mel Frequency Cepstral Coefficient (MFCC) and Dynamic Time Warping (DTW) Techniques (2010, methodology): “Using dynamic programming techniques, the search for the minimum distance path can be done in polynomial time P(t), using equation below[14]: where, N is the length of the sequence, and V is the num ‐ ber of templates to be considered.”
Three Myths about Dynamic Time Warping Data Mining (2005, methodology): “who work with chromatographic data noted “Unconstrained dynamic time warping was found to be too flexible for this chromatographic data set, resulting in a overcompensation of the observed shifts” [20], or Rath & Manmatha have carefully optimized the constraints for the task of indexing historical a”
Indexing multi-dimensional time-series with support for multiple distance measures (2003, methodology): “It has been used for tracking time series extracted from video [7], classifying handwritten text [16] and even fingerprint indexing [13].”
Recognition and retrieval of mathematical expressions (2011, methodology): “Connected components in the query image are matched with connected components in archived images using visual similarity of connected components, again based on features similar to Rath and Manmatha's.”
Whole is Greater than Sum of Parts: Recognizing Scene Text Words (2013, methodology): “(1) Our matching score is based on a novel feature set, which shows better performance than the profile features in [16].”
Semi-supervised time series classification (2006, methodology): “We test on the Word Spotting dataset, which was created by Rath and Manmatha for word image matching [36].”