3 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 3 / 3
Yayın Biometric recognition using bio-signals(Işık Üniversitesi, 2017-04-14) Dursun, Ceren; Gürkan, Hakan; Güz, Ümit; Işık Üniversitesi, Fen Bilimleri Enstitüsü, Elektronik Mühendisliği Yüksek Lisans ProgramıThe main objective of the project is to increase the recognition rate by establishing a multimodal biometric recognition system that uses two di_erent biometric characteristics, as bio-signals. Today, institutions use biometric recognition systems quite often to provide security for many areas such as information security and physical security. The importance of these systems increases day by day in the direction of technological development and increasing demand. Recognition systems based on biometric characteristics are more reliable, because of the possibility of forgetting or losing knowledge in the recognition systems based on knowledge (eg: password) and the possibility of being stolen or guessed by third persons in the biometric recognition based on possessed (eg: card). However, the fraud techniques are also developing in the direction of technological developments and biometric characteristics cannot be renewed in case of imitation, hence the use of multiple biometrics recognition system may be a solution to this problem. At the same time, the use of multiple biometrics increases in the security of systems. In this thesis, a biometric recognition system, which uses the lectrocardiogram (ECG) and speech signals of the person, was created. Since there was not enough time and possibility, an arti_cial database was generated with obtaining these signals from various sources. First, the MIT-BIH Arrhythmia Database was used for ECG signals. This database consists of 48 ECG signals, which belong to 22 females and 26 males. In ccordance with this database, a database was created for the speech signals, which were obtained from the website, given in [1]. The features of the biometric signals were extracted by AC / DCT (Autocorrelation/ Discrete Cosine Transform) method for ECG signals and by Mel Frequency Cepstrum Coe_cients (MFCCs) method for speech signals. The data, which were obtained from the feature extraction, were then classi_ed by the Gaussian Mixture Model (GMM) method. The scores, which were obtained from the classi_cation process, were fused as a single individual's data, and the decision-making step was passed. Recognition rates were obtained in the decision making step. The recognition rate for the ECG signal was %87.50 and 42 persons were matched correctly. The recognition rate for 2 seconds speech signals was %58.33 and 28 persons were matched correctly. Normalization was applied before the fusion of these two datasets. The recognition rate after the fusion was %70.83 and 34 persons were matched correctly. However, when the recognition rates are considered, it has been observed that the recognition rate, which obtained after the fusion, is lower than recognition rate of the ECG signals. Therefore, instead of 2 seconds speech signals, 10 seconds speech signals were used. In this case, the recognition rate of the speech signals was %97.9 and 47 persons were matched correctly. Then, normalization was applied again and two datasets were fused. After the fusion, the rate of recognition reached %95.8 and 46 persons were matched correctly.Yayın Automatic speech recognition system for Turkish spoken language(Işık Üniversitesi, 2012-06-21) Dalva, Doğan; Güz, Ümit; Işık Üniversitesi, Fen Bilimleri Enstitüsü, Elektronik Mühendisliği Yüksek Lisans ProgramıThe transmission and storage of speech sounds is possible for decades. In addition by using signal processing techniques, it is also possible tp process speech signals. By using time abd frequency analysis od speech signal and several machine learning algorithms, it is possible to build a system which is used to recognize spoken words. Such systems are called Automatic Speech Recognition systems. In our work, We have used the Automatic Speech Recognition system for Turkish spoken language which has built by BUSIM speech group. However, the output of the recognizer is the list of spoken words. Even for humans it is avery hard to understand a text without punctuation symbols. Hence to build more complex recognizer whose goal to perform topic segmentation and topic summarization, the output of ASR should be divided into sentences at first. Our goal is to build a system which performs the sentence segmentation. In our work We have used ASR system to obtain word level and phoneme level time marks and by using that time marks with the audio files, We have extracted prosodic features, where the prosodic properties of speech contains information about the punctuation in the text, which is not available at the output of ASR system.Yayın Prosodic, morphological and lexical feature extraction of Turkish broadcast news data(Işık Üniversitesi, 2014-06-05) Revidi, İzel D.; Güz, Ümit; Işık Üniversitesi, Fen Bilimleri Enstitüsü, Elektronik Mühendisliği Yüksek Lisans ProgramıSentence segmentation from speech is part of a process that aims at enriching the unstructured stream of words that are the output of standard speech recognizers. Its role is to find the sentence units in this stream of words. Sentence segmentation is a preliminary step toward speech understanding. Once the sentence boundaries are detected, further syntactic and/or semantic analysis can be performed on these sentences. Usually, speech recognizer output lacks the textual cues to these entities (such as headers, paragraphs, sentence punctuation, and capitalization). However, speech provides extra non-lexical cues, related to features like pitch, energy, pause and word durations as prosodic features; verb, noun or adjective as a morphological features and also lexical features. These prosodic, morphological and lexical features are provides a complementary information for segmentation of speech into sentences. Our goal is examine feature the extraction and use of prosodic information which has been done in previous works, in addition to lexical features and morphological for spoken language processing of Turkish with open source tools.
