Conference Papers

Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506

Browse

Subject: search.filters.subject.HMM

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Offline signature verification based on contourlet transform and textural features using HMM
    (Institute of Electrical and Electronics Engineers Inc., 2014) Pushpalatha, K.N.; Supreeth Prajwal, S.S.; Gautam, A.K.; Shiva Kumar, K.B.S.
    Automatic offline signature verification and recognition is becoming essential in personal authentication. In this paper, we propose a transform domain offline signature verification system based on contourlet transform, directional features and Hidden Markov Model (HMM) as classifier. The signature image is preprocessed for noise removal and a two level contourlet transform is applied to get feature vector. The textural features are computed and concatenated with coefficients of contourlet transform to form the final feature vector. HTK tool with HMM classifier is used for classification. The parameters of False Rejection Rate (FRR), False Acceptance Rate (FAR) and Total Success Rate (TSR) are calculated for GPDS-960 database. It is found that the parameters of FRR and FAR are improved compared to the existing algorithms. © 2014 IEEE.
  • No Thumbnail Available
    Item
    Monophone and Triphone Acoustic Phonetic Model for Kannada Speech Recognition System
    (Institute of Electrical and Electronics Engineers Inc., 2022) Kumar, T.N.M.; Jayan, A.; Bhat, S.; Anvith, M.; Narasimhadhan, A.V.
    The automatic Speech Recognition system (ASR) is the most widely used application in the speech domain. ASR systems generate text data from spoken utterances without manual intervention. In this work, we build an ASR system for the Kannada language. For building the proposed system, we extract Mel Frequency Cepstral Coefficients (MFCC) features from the audio data, and the Kannada language model is developed using corresponding labels. The dictionary generation and phonetic labelings are automated. Recognition performance is compared for both monophonic and triphone models. The word error rate of 15.73 % and the sentence error rate of 55.5 % are achieved for the triphone model. Comparatively, the triphone model gives a better performance than the monophonic model. © 2022 IEEE.