Text Dependent Writer Verification using Boosting
نویسندگان
چکیده
Text-dependent writer verification systems are preferred over text-independent systems due to the accuracy they achieve with small amount of data. However, text-dependent systems are prone to forgery. This paper proposes a novel boosting based framework for writerspecific text generation to increase the accuracy and a method of text variation to make the system robust to forgery. The approach is able to achieve error rates of 5% with just 6 words as compared to random(11%) or most discriminative(22%) primitive selection methods on a dataset containing 20 writers. Boosting based text selection also provides the flexibility to incorporate text variation across multiple authentications, which in turn makes the system robust to forgery.
منابع مشابه
A Proposal for a Text-Indicated Writer Verification Method - Document Analysis and Recognition, 1997., Proceedings of the Fourth International Conference on
W e propose an on-line writer verification method to improve the reliability of verifying a specific system user. In the proposed method, a different text including ordina y characters is used on eve y verification process. This text can be selected automatically by the verification system so as to reflect the specific writer's fentures. A specific writer is accepted only when the same text, wh...
متن کاملWriter Demographic Classification Using Bagging and Boosting
Classifying handwriting into a writer demographic category, e.g., gender, age, or handedness of the writer, is useful for more detailed analysis such as writer verification and identification. This paper describes classification into binary demographic categories using document macro features and several different classification methods: a single feed-forward neural network classifier and combi...
متن کاملState of the art in off-line writer identification of handwritten text and survey of writer identification of Arabic text
In this paper we present the state of the art in writer identification and verification of handwritten text. In addition, a special and extensive survey of writer identification and verification of Arabic handwritten text is also included. Different feature extraction techniques are addressed showing the different research groups’ efforts as well as individual efforts. The different classificat...
متن کاملWriter Verification by Using an Arbitrary Part of Feature Sequence Extracted from On-Line Handwritten Character/Figure Patterns Hiroshi Kameya A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHIROSOPHY IN COMPUTERSCIENCE AND ENGINEERING
Writer Verification by Using an Arbitrary Part of Feature Sequence Extracted from On-Line Handwritten Character/Figure Patterns Hiroshi Kameya Chair of Supervisory Committee: Professor Ryuichi Oka Graduate Department of Information Systems By introducing the continuous dynamic programming (CDP) algorithm developed by Oka (1998), I have developed a new segmentation-free, text-independent biometr...
متن کاملText-indicated Writer Verification Using Hidden Markov Models
We propose an HMM-based text-indicated writer verification method, which is based on a challenge and response type of authentication process. In this method, a different text including ordinary characters is used on every occasion of verification. This text can be selected automatically by the verification system so as to reflect a specific writer's personal features. The specific writer is acc...
متن کامل