Hierarchical Dictionary Learning and Sparse Coding for Static Signature Verification

An assortment of review papers as well as newly quoted literature indicates that usually, the most important link in the chain of designing signature verification systems (SV's) is the feature extraction one. These methods are divided in two main categories. The first one, includes handcrafted features which are methods smanually engineered by scientists to be optimal for certain type of information extraction-summarization from signature images. Examples of this kind include global-local and/or grid-texture oriented features. The second feature category addresses signature modeling and verification with the use of dedicated features, usually learned directly from raw signature image data. Typical representatives include Deep Learning (DL) as well as Bag of Visual Words (BoW) or Histogram of Templates (HOT). Recently, sparse representation (SR) methods (dictionary learning and coding) have been introduced for signature modeling and verification with promising results. In this paper, we propose an extension of the SR framework by introducing the idea of embedding the atoms of a dictionary in a directed tree. This is demonstrated with an l
₀
tree-structured sparse regularization norm. The efficiency of the proposed method is shown by conducting experiments with two popular datasets namely the CEDAR and MCYT-75.