Online Writer Identification Using Sparse Coding and Histogram Based Descriptors

Dwivedi, Isht; Gupta, Shiv Narain; Venugopal, Vivek; Sundaram, Suresh

doi:10.1109/icfhr.2016.0110

Cited by 9 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The The bold values denote that the method achieve the best performance The bold values denote that the method achieve the best performance writing trajectory. For example, the writing examples of IDs "28" and "11" in the ranking list at the first row are very similar, but they can be distinguished from the distinctive online segments of the letter "g" (Color figure online) (Dwivedi 2016) under the open-set setting on LERID\ Modified IAM (IAM). The results imply that the letter-level online writer-id is challenging.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…For the closed-set setting, we compare our model with two hand-crafted models (point-based (Gargouri et al 2013) and histogram-based (Dwivedi 2016)) and two deep models (DeepWriterID (Yang et al 2016) and DeepRNN (Zhang et al 2017)) on three online writer-id datasets (Modified IAM (IAM), LetWriterDB (Chen et al 2018) and LERID). The The bold values denote that the method achieve the best performance The bold values denote that the method achieve the best performance writing trajectory.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…Conventionally, most researches on writer-id are focused on document-level and text-level analysis. Some methods (Tsai and Lan 2005;Namboodiri and Gupta 2006;Bulacu and Schomaker 2007;Schlapbach and Bunke 2007;Li et al 2007;Schlapbach et al 2008;El Abed et al 2009;Tan et al 2010;Chaabouni et al 2011;Ramaiah et al 2013;Shivram et al 2013;Bertolini et al 2013;Dwivedi 2016;Lai and Jin 2019;Khan et al 2018;Nguyen et al 2019) exploit sophisticated hand-crafted features to model handwriting. For example, El et al (El Abed et al 2009) modelled writing with temporal sequences and shape codes; modelled writing based on a codebook extracted by clustering the point and sub-stroke descriptions; (Lai and Jin 2019) modelled writing based on path signature features (Chen 1958); and (Khan et al 2018) modelled writing with visual descriptors (e.g., SIFT (Arandjelović and Zisserman 2012)).…”

Section: Writer Identificationmentioning

confidence: 99%

“…However, existing writer-id methods (Namboodiri and Gupta 2006;Gargouri et al 2013;Dwivedi 2016;Li and Tan 2009;Singh and Sundaram 2015;Venugopal and Sundaram 2017;Li et al 2007) are restricted to text-\ document-level analysis (i.e., they require a long text or a full document for identification), which limits their scalability and flexibility and thus hinders their practical application. For example, on a smartphone, it is very inconvenient to write a full document.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Letter-Level Online Writer Identification

Chen

et al. 2021

Int J Comput Vis

View full text Add to dashboard Cite

Writer identification (writer-id), an important field in biometrics, aims to identify a writer by their handwriting. Identification in existing writer-id studies requires a complete document or text, limiting the scalability and flexibility of writer-id in realistic applications. To make the application of writer-id more practical (e.g., on mobile devices), we focus on a novel problem, letter-level online writer-id, which requires only a few trajectories of written letters as identification cues. Unlike text-\ document-based writer-id which has rich context for identification, there are much fewer clues to recognize an author from only a few single letters. A main challenge is that a person often writes a letter in different styles from time to time. We refer to this problem as the variance of online writing styles (Var-O-Styles). We address the Var-O-Styles in a capture-normalize-aggregate fashion: Firstly, we extract different features of a letter trajectory by a carefully designed multi-branch encoder, in an attempt to capture different online writing styles. Then we convert all these style features to a reference style feature domain by a novel normalization layer. Finally, we aggregate the normalized features by a hierarchical attention pooling (HAP), which fuses all the input letters with multiple writing styles into a compact feature vector. In addition, we also contribute a large-scale LEtter-level online wRiter IDentification dataset (LERID) for evaluation. Extensive comparative experiments demonstrate the effectiveness of the proposed framework.

show abstract

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Writer Identificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Letter-Level Online Writer Identification

Chen

et al. 2021

Int J Comput Vis

View full text Add to dashboard Cite

show abstract

Online writer identification system using adaptive sparse representation framework

Venugopal

Sundaram

2020

IET biom.

Self Cite

View full text Add to dashboard Cite

This study explores an adaptive sparse representation approach for online writer identification. The main focus is on employing prior information that quantifies the degree of importance of a dictionary atom concerning a given writer. This information is proposed by a fusion of two derived components. The first component is a saliency measure obtained from the sum-pooled sparse coefficients corresponding to the sub-strokes of a set of enrolled writers. The second component is a similarity score, computed for each dictionary atom with regards to a given writer, that is related to the reconstruction error of the sub-stroke based feature vectors. The proposed identification is accomplished with an ensemble of support vector machines (SVMs), wherein the input to the SVM trained for a writer is obtained by incorporating the adapted saliency values of that writer on the document descriptor obtained via average pooling of sparse codes. Experiments performed on the IAM and IBM-UB1 online handwriting databases demonstrate the efficacy of the proposed scheme.

show abstract

Letter-Level Writer Identification

Chen

et al. 2018

2018 13th IEEE International Conference on Automatic Face &Amp; Gesture Recognition (FG 2018)

View full text Add to dashboard Cite

Online Writer Identification Using Sparse Coding and Histogram Based Descriptors

Cited by 9 publications

References 14 publications

Letter-Level Online Writer Identification

Letter-Level Online Writer Identification

Online writer identification system using adaptive sparse representation framework

Letter-Level Writer Identification

Contact Info

Product

Resources

About