Signature recognition

About: Signature recognition is a research topic. Over the lifetime, 2138 publications have been published within this topic receiving 37605 citations.

Combination of three classifiers with different architectures for handwritten word recognition

Abstract: The study of multiple classifier systems has become an area of intensive research in pattern recognition recently. Also in handwriting recognition, systems combining several classifiers have been investigated. In this paper the combination of three classifiers for handwritten word recognition with different architectures is studied. In addition a new ensemble method working with several base classifiers is applied and the results of the ensemble method are compared to the results of the combination of the three classifiers. In the experiments a large-scale handwritten word recognition task is considered.

Development of Biometric Systems for Person Recognition: Biometric Feature Systems, Traits and Acquisition

Abstract: In this paper biometric feature systems and multimodal databases for biometric recognition systems are analyzed on the basis of scientific publications in IEEE Xplore digital library. It is shown that wavelet transform coefficients are the most universal feature used in biometric person recognition systems – it is among five frequently used features used in all five popular traits. Moreover, face is the most frequently used trait in multimodal person recognition systems – it is used along with 48 % of iris, 44 % of fingerprint, 33 % of voice and 24 % of signature multimodal systems. Analysis of 15 multimodal databases reveals the fact, that older multimodal databases, e. g., XM2VTS, are still widely used for comparison. However databases such as Biosecure are more versatile and should become more popular soon if a free access will be provided.

System and method for hierarchical cryptographic key generation using biometric data

Abstract: A cryptographic key is generated using biometric data and a hierarchy of biometric descriptors. The hierarchy of biometric descriptors includes multiple levels, wherein a biometric descriptor at a first level is associated with a subset of the biometric descriptors at the next lower level. To generate a cryptographic key, biometric data is collected and compared to the biometric descriptors at the first level of the hierarchy. One of the biometric descriptors is selected at the first level, and a first key component is generated based on the first selected biometric descriptor. The biometric data is then compared to the subset of biometric descriptors at the second level of the hierarchy associated with the first selected biometric descriptor. This process of selecting a biometric descriptor and generating a key component continues for each level of the hierarchy. The key components are then used to generate a cryptographic key.

Face recognition using improved principal component analysis

Abstract: At the general recognition process, the feature vectors that are obtained from some facial images are transformed into recognition space by Fisher's linear discriminate method (Fisher's method) and principal component analysis (PCA). But at Fisher's method we must recalculate all recognition space when adding a registrant or registrant's learning patterns. In contrast, though at PCA we only recalculate added registrant's pace when adding, the face recognition rate obtained from the conventional PCA is bad, because the aim of the conventional PCA is dimension curtailment for compression of data and isn't dimension curtailment for recognition. Therefore we proposed improved principal component analysis (IPCA) for pattern recognition.

Probabilistic vector mapping with trajectory information for noise-robust speech recognition

Abstract: The authors propose a probabilistic vector mapping method with trajectory information for noise-robust speech recognition. The proposed method estimates a speech vector using not only the current noisy vector but also adjacent vectors in the form of a trajectory pattern. According to our isolated word recognition experiment, the new method yields a better recognition rate than the probabilistic vector mapping method previously proposed.