IJIGSP Vol. 5, No. 9, Jul. 2013
Cover page and Table of Contents: PDF (size: 137KB)
Acoustic analysis is a proper method in vocal fold pathology diagnosis so that it can complement and in some cases replace the other invasive, based on direct vocal fold observation, methods. There are different approaches and algorithms for vocal fold pathology diagnosis. These algorithms usually have three stages which are Feature Extraction, Feature Reduction and Classification. While the third stage implies a choice of a variety of machine learning methods (Support Vector Machines, Artificial Neural Networks, etc), the first and second stages play a critical role in performance and accuracy of the classification system. In this paper we present initial study of feature extraction and feature reduction in the task of vocal fold pathology diagnosis. A new type of feature vector, based on wavelet packet decomposition and Mel-Frequency-Cepstral-Coefficients (MFCCs), is proposed. Also a new GA-based method for feature reduction stage is proposed and compared with conventional methods such as Principal Component Analysis (PCA). Support vector machine is used as a classifier for evaluating the performance of the proposed method. The results show the priority of the proposed method in comparison with the current methods.[...] Read more.
The purpose of speech emotion recognition system is to classify speaker's utterances into different emotional states such as disgust, boredom, sadness, neutral and happiness.
Speech features that are commonly used in speech emotion recognition (SER) rely on global utterance level prosodic features. In our work, we evaluate the impact of frame-level feature extraction. The speech samples are from Berlin emotional database and the features extracted from these utterances are energy, different variant of mel frequency cepstrum coefficients (MFCC), velocity and acceleration features. The idea is to explore the successful approach in the literature of speaker recognition GMM-UBM to handle with emotion identification tasks. In addition, we propose a classiﬁcation scheme for the labeling of emotions on a continuous dimensional-based approach.
In this paper we demonstrate the impact of language parameter variability on mono, cross and multi-lingual speaker identification under limited data condition. The languages considered for the study are English, Hindi and Kannada. The speaker specific features are extracted using multi-taper mel-frequency cepstral coefficients (MFCC) and speaker models are built using Gaussian mixture model (GMM)-universal background model (UBM). The sine-weighted cepstrum estimators (SWCE) with 6 tapers are considered for multi-taper MFCC feature extraction. The mono and cross-lingual experimental results show that the performance of speaker identification trained and/or tested with Kannada language is decreased as compared to other languages. It was observed that a database free from ottakshara, arka and anukaranavyayagalu results a good performance and is almost equal to other languages.[...] Read more.
Edge detection is the most fundamental but at the same time most important task in image processing and analysis. In the paper a hybrid approach combining Neural Network and Fuzzy logic based edge detection algorithm is proposed to detect edges in grayscale images. To improve the generalization ability, the neural network is trained on fuzzy inputs rather than crisp inputs. The network consists of three layers, one input layer, one hidden layer and one output layer. Fuzzy membership functions are used to convert neurons of input and hidden layer into fuzzy neurons. So the output of first and second layer is the membership value of the corresponding input in the fuzzy set. The proposed technique provides advantage of both neural networks and fuzzy logic and gives satisfactory results for both noisy and noise free images. The method is compared with Roberts, Prewitt, Sobel and Laplacian of Gaussian and other neural network and fuzzy logic based methods and the experimental results reveal that proposed method gives better edge map considering the problem of false edge detection.[...] Read more.
Recognizing a plant in any huge vegetation is a tedious work for us. We recognize a plant on the basis of its size, leaves, flowers, fruits, etc. Leaf is a part of the plant which can be found on plants almost in all seasons and most of the time we have to recognize plants on the basis of its leaf. But when dealing with leaf of plant, it is important to consider the finer details of the contour representing the shape of the leaf. We are trying to build a system which has a database of leaves of different plants and given a leaf, we find out the plant to which it may belong. In this paper, we present the results of tangential angle approach used for retrieval. A database of around one thousand leaves of different plants has been created. Each leaf image is preprocessed to extract its boundary. Then tangential angle approach is applied which captures the angular details of the boundary of shape. We have done the testing for around 1000 leaves and on the basis of that recall, precision and error rate have been calculated to measure the effectiveness of the proposed method.[...] Read more.
Most image processing algorithms are inherently parallel, so multithreading processors are suitable in such applications. In huge image databases, image processing takes very long time for run on a single core processor because of single thread execution of algorithms. Graphical Processors Units (GPU) is more common in most image processing applications due to multithread execution of algorithms, programmability and low cost. In this paper we implement texture based image retrieval system in parallel using Compute Unified Device Architecture (CUDA) programming model to run on GPU. The main goal of this research work is to parallelize the process of texture based image retrieval through entropy, standard deviation, and local range, also whole process is much faster than normal. Our work uses extensive usage of highly multithreaded architecture of multi-cored GPU. We evaluated the retrieval of the proposed technique using Recall, Precision, and Average Precision measures. Experimental results showed that parallel implementation led to an average speed up of 140.046×over the serial implementation. The average Precision and the average Recall of presented method are 39.67% and 55.00% respectively.[...] Read more.
This paper presents a new approach for designing a speaker recognition system based on mel frequency cepstral coefficients (MFCCs) and voice activity detector (VAD). VAD has been employed to suppress the background noise and distinguish between silence and voice activity. MFCCs were extracted from the detected voice sample and are compared with the database for recognition of the speaker. A new criteria for detection is proposed which gives very good performance in noisy environment.[...] Read more.
The process of recognition and identification of plant species is very time-consuming as it has been mainly carried out by botanists. The focus of computerized living plant's identification is on stable feature's extraction of plants. Leaf-based features are preferred over fruits, also the long period of its existence than fruits. In this preliminary study, we study and propose neural networks and Mutual information for identification of two, three Fig cultivars (Ficus Carica L.) in Syria region. The identification depends on image features of Fig tree leaves. A feature extractor is designed based on Mutual Information computation. The Neural Networks is used with two hidden layers and one output layer with 3 nodes that correspond to varieties (classes) of FIG leaves. The proposal technique is a tester on a database of 84 images leaves with 28 images for each variety (class). The result shows that our technique is promising, where the recognition rates 100%, and 92% for the training and testing respectively for the two cultivars with 100% and 90 for the three cultivars. The preliminary results obtained indicated the technical feasibility of the proposed method, which will be applied for more than 80 varieties existent in Syria.[...] Read more.