Browsing by Author "Djoudjai, Mohamed Anis"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Feature Selection using F-score Method for Offline Arabic Handwritten Fragment Identification
    (University of Eloued جامعة الوادي, 2022-01-24) Azzoug, Soraya; Chibani, Youcef; Djoudjai, Mohamed Anis
    Among the various means of document authentication, handwriting fragments represent one of the challenging tasks for automatic handwritten writer identification. Usually, the writer is represented by a set of features extracted for each handwritten fragment. However, feature components might be irrelevant due to their redundancy, affecting the pertinence of the feature vector. Hence, this paper proposes a feature-selection strategy based on a hybrid F-score method, performed genuinely for multi-class classification. The F-score is used conjointly to a classifier based-distance. According to a statistical analysis performed on F-score distribution during the training step, the number of pertinent feature components is deduced from the highest F-score value. The experimental evaluation performed on the well-known IFN/ENIT handwritten fragment dataset shows an improvement of the identification rate of 94.20% while reducing the size of the feature vector of 54%.
  • No Thumbnail Available
    Item
    Interval versus Histogram of Symbolic Representation Based One-Class Classifier for Offline Handwritten Signature Verification
    (University of Eloued جامعة الوادي, 2022-01-24) Djoudjai, Mohamed Anis; Chibani, Youcef
    This paper proposes a comparison study of using Interval and Histogram of Symbolic Representation (ISR and HSR) based One-Class classifiers, namely OC-ISR and OC-HSR, respectively, applied to the offline signature verification. Usually, symbolic verification models are built straightforward from the feature space. The proposed work explores an alternative approach based on the use of feature-dissimilarities generated from Curvelet Transform (CT) for building the OC-ISR and the OC-HSR classifier. For the OC-ISR classifier, a new weighted membership function is proposed for computing the similarity values between a dissimilarity query vector and a targeted ISR model. The experimental evaluation performed on the well-known public datasets GPDS, CEDAR, and MCYT, reveals the proposed OC-ISR's superiority over the OC-HSR classifier. Moreover, the proposed verification model based on the OC-ISR classifier outperforms the last similar work reported in the literature on the GPDS-160 dataset by 0.99%, 0.8%, and 0.35% of Average Error Rate (AER) for 5, 8, and 12 reference signatures, respectively.