Mohd. Rizon Mohamed Juhari, Prof. Ir. Dr.http://dspace.unimap.edu.my:80/xmlui/handle/123456789/229582024-03-29T05:54:03Z2024-03-29T05:54:03ZIdentification of vocal and voice disordersMurugesa Pandiyan, Paulraj, Prof. Madya Dr.Sazali, Yaacob, Prof. Dr.Mohd Rizon, Mohammed Juhari, Prof. Dr.Sivanandam, S. N.Muthusamy, Hariharan, Dr.http://dspace.unimap.edu.my:80/xmlui/handle/123456789/151282011-10-28T08:18:52Z2007-10-25T00:00:00ZIdentification of vocal and voice disorders
Murugesa Pandiyan, Paulraj, Prof. Madya Dr.; Sazali, Yaacob, Prof. Dr.; Mohd Rizon, Mohammed Juhari, Prof. Dr.; Sivanandam, S. N.; Muthusamy, Hariharan, Dr.
The discrimination of normal and pathological voices using noninvasive acoustic analysis helps to perform accurate identification of voice disorders and diagnoses of vocal and voice disease. Acoustic analysis is a non- invasive technique based on digital processing of the speech signal. In the recent years, acoustic analysis of normal and pathological voices have become increasingly interesting to researchers in laryngology and speech pathologies. This paper presents classification of pathological voices using neural network trained by Back propagation algorithm with slope parameter and BP with binary sigmoidal and Gaussian activation function. Simulation results indicate that the proposed algorithm provide better classification rate than conventional back propagation algorithm.
Organized by Universiti Malaysia Perlis (UniMAP), 25th - 26th October 2007 at Putra Brasmana Hotel, Kuala Perlis, Perlis, Malaysia.
2007-10-25T00:00:00ZFacial features for template matching based face recognitionChai, Tong YuenMohd Rizon, Mohamed Juhari, Prof. Dr.Woo, San SanTan, Ching Seong, Dr.http://dspace.unimap.edu.my:80/xmlui/handle/123456789/150902011-10-27T07:02:23Z2009-01-01T00:00:00ZFacial features for template matching based face recognition
Chai, Tong Yuen; Mohd Rizon, Mohamed Juhari, Prof. Dr.; Woo, San San; Tan, Ching Seong, Dr.
Problem statement: Template matching had been a conventional method for object detection especially facial features detection at the early stage of face recognition research. The appearance of moustache and beard had affected the performance of features detection and face recognition system since ages ago. Approach: The proposed algorithm aimed to reduce the effect of beard and moustache for facial features detection and introduce facial features based template matching as the classification method. An automated algorithm for face recognition system based on detected facial features, iris and mouth had been developed. First, the face region was located using skin color information. Next, the algorithm computed the costs for each pair of iris candidates from intensity valleys as references for iris selection. As for mouth detection, color space method was used to allocate lips region, image processing methods to eliminate unwanted noises and corner detection technique to refine the exact location of mouth. Finally, template matching was used to classify faces based on the extracted features. Results: The proposed method had shown a better features detection rate (iris = 93.06%, mouth = 95.83%) than conventional method. Template matching had achieved a recognition rate of 86.11% with acceptable processing time (0.36 sec). Conclusion: The results indicate that the elimination of moustache and beard has not affected the performance of facial features detection. The proposed features based template matching has significantly improved the processing time of this method in face recognition research.
Link to publisher's homepage at http://thescipub.com
2009-01-01T00:00:00ZAutomated face localization and facial features detection using geometric informationChai, Tong YuenMohd Rizon, Mohamed Juhari, Prof. Dr.Karthigayan, M.Sugisaka, M.Hazry, Desa, Dr.Ibrahim, Z.http://dspace.unimap.edu.my:80/xmlui/handle/123456789/150622011-10-27T04:42:02Z2007-11-12T00:00:00ZAutomated face localization and facial features detection using geometric information
Chai, Tong Yuen; Mohd Rizon, Mohamed Juhari, Prof. Dr.; Karthigayan, M.; Sugisaka, M.; Hazry, Desa, Dr.; Ibrahim, Z.
This paper presents algorithms to detect irises of both eyes and mouth from color and intensity images and extracts intensity valleys from the race region. Next, the algorithm extracts iris candidates from the valleys and computes the costs for each of iris candidates. Finally, a pair if iris candidates is selected as irises by using the computed costs. Projection based method has been used to detect mouth corresponding to irises location. This algorithm has been tested with images from European database and South East Asian database.
Organized by Malaysia-Japan University Center (MJUC), 12th - 15th November 2007 at Seri Pacific Hotel, Kuala Lumpur, Malaysia.
2007-11-12T00:00:00ZApplication of feedforward neural network for the classification of pathological voicesSazali, Yaacob, Prof. Dr.Murugesa Padiyan, Paulraj, Dr.Mohd Rizon, Mohammed Juhari, Prof. Dr.Muthusamy, Hariharan, Dr.http://dspace.unimap.edu.my:80/xmlui/handle/123456789/148752011-10-24T03:53:00Z2007-03-09T00:00:00ZApplication of feedforward neural network for the classification of pathological voices
Sazali, Yaacob, Prof. Dr.; Murugesa Padiyan, Paulraj, Dr.; Mohd Rizon, Mohammed Juhari, Prof. Dr.; Muthusamy, Hariharan, Dr.
This paper present the application of feed forward neural network for the classification of pathological voices based on the on the acoustic analysis and EGG features. Acoustic analysis is a non-invasive technique based on digital processing of the speech signal. Electroglottography is a method of obtaining vibration signal related to the laryngeal phonatory function. The Electroglottograph (EGG) is an instrument that register the contact between the vocal folds as a time-varying signal. The time domain voice parameters are computed from the extracted pitch data. In this paper, a Feedback Neural Network is employed for the classification of pathological voices. The Acoustic parameters extracted from the speech signal and the features from the Electroglottography from the input to the neural network distinguish the voice as pathological or a non-pathological voice.
Organized by Universiti Teknologi MARA, 9th - 11th March 2011 at Malacca, Malaysia.
2007-03-09T00:00:00Z