Publications


35 Records Found

Abstract: Ability of recognizing facial expression is important part of behavioural science, which helps to ease the communication. This ability can serve in many contexts. Hence, facial expression is an important research area over the last two decades. In this paper, describes the extraction of the minimum number of Gabor wavelet parameters for the recognition of facial expressions and work with facial components like eyes and mouth by using hierarchical approach. The objective of our research was to investigate the performance of a facial expression recognition system and less work with feature extraction to classify expression. This system recognizes basic seven expressions happy, sad, neutral, angry, surprise, fear and disgust. We present a hierarchy for facial region extraction from static image. For determination of face effective areas is used from bounding box. This method has high ability in intelligent selection of areas in facial expression recognition system. Using determination of effective areas classify expression directly. Remaining faces fed to Gabor filter and it further reduces by Principle Component Analysis (PCA) to classify expression using Euclidean distance. Results test on JAFFE database indicates that proposed system for facial expression recognition is good accuracy and generating superior results as compared to other approaches.

Index Terms: Facial Expression Recognition, Feature Extraction, Gabor Filter, Principle Component Analysis, Linear Discriminant Analysis

Abstract: Sorting is the important operation and well-studied problem in computer science. Sorting refers to the operation of arranging data in some given order such as ascending or descending, with numerical data, or lexicographical order. Data being sorted may be single valued data or it may be satelight data. So according to the cost of memory and write operations, sorting technique must be chosen. Sorting algorithms are compared based on complexity (number of comparison, number of swaps etc), methods used like comparison-based or non-comparison based, internal sorting or external sorting etc. There are many sorting algorithm that have been proposed to meet the particular application. This paper shows a way to improve the performance of traditional selection sort algorithm. Results shows that the proposed approach outperforms the traditional selection sort algorithm in terms of number of comparisons.

Index Terms: Sorting, Complexity, Selection sort, Insertion sort, Bubble sort.

Abstract: Character recognition comes into picture when various patterns of handwritten or optical characters are to be recognized digitally. Many researchers have proposed different approaches for character recognition in different languages. In this paper, we have reviewed several techniques of character recognition. The main important phases of character recognition include preprocessing, segmentation, feature extraction and classification. Various feature extraction techniques and classification techniques have been surveyed in this paper and an attempt is made here to draw a conclusion regarding HCR techniques from the literature survey.

Index Terms: HCR, Segmentation, Neural Network, Feature Extraction, Classification

Abstract: Content based image retrieval (CBIR) considers the characteristics of the image itself, for example its shapes, colors and textures. CBIR has many applications areas such as, education, commerce, military, searching, bio medicine and web image classification. The domain of CBIR is expanding day by day, the requirements have become complex and so are the algorithms. CBIR is a new but widely adopted method for finding images from vast and unannotated image databases. In this paper we will discuss a technique known as Multistage CBIR. The proposed technique consists of a three layer feed forward architecture i.e. the first layer consists of comparison of color features the second consist of comparing texture feature and the last is comparing shape features.

Index Terms: CBIR; Content Based Image Retrieval; Multistage Content Based Image Retrieval; Color Feature Extraction; Texture Feature Extraction; Shape Feature Extraction

Abstract: Facial Expression Recognition (FER) is a rapidly growing and ever green research field in the area of Computer Vision, Artificial Intelligent and Automation. There are many applications which use Facial Expression to evaluate human nature, feelings, judgment, and opinion. Recognizing Human Facial Expression is not a simple task because of some circumstances due to illumination, facial occlusions, face color/shape etc. In these paper, we present some method/techniques such as Principal Component Analysis (PCA), Linear Discriminate Analysis (LDA), Gabor Filter/Energy, Line Edge Mapping (LEM), Neural Network, Independent Component Analysis (ICA) which will directly or/and indirectly used to recognize human expression in various condition. In this paper, we proposed Gabor Filter which is one of the effective methods to extract facial features from digital images. Gabor Filter is biological relevant with human virtual cortex and computational properties. To evaluate recognition rate and accuracy, we have used Japanese Female Facial Expression Database (JAFFE). Some experiments, whose satisfactory results prove the effectiveness of this recognition approach are also illustrate here.

Index Terms: Image Processing, Facial Expression Recognition, Gabor Filter, Gaussian kernel, Wavelength

Abstract: Content based image retrieval (CBIR) considers the characteristics of the image itself, for example its shapes, colors and textures. CBIR has many applications areas such as, education, commerce, military, searching, bio medicine and web image classification. The domain of CBIR is expanding day by day, the requirements have become complex and so are the algorithms. CBIR is a new but widely adopted method for finding images from vast and unannotated image databases. In CBIR images are indexed on the basis of low-level features, such as color, texture, and shape that can automatically be derived from the visual content of the images. In this paper we have discussed five different image processing techniques.

Index Terms: CBIR, Content Based Image Retrieval, Color histogram, Object Detection, Comparison of CBIR, Multistage Content Based Image Retrieval

Abstract: Facial Expression Recognition (FER) is a rapidly growing and ever green research field in the area of Computer Vision, Artificial Intelligent and Automation. There are many application which uses Facial Expression to evaluate human nature, feelings, judgment, opinion. Recognizing Human Facial Expression is not a simple task because of some circumstances due to illumination, facial occlusions, face color/shape etc. In these paper, we present some method/techniques such as Principal Component Analysis (PCA), Linear Discriminate Analysis (LDA), Gabor Filter/Energy, Line Edge Mapping (LEM), Neural Network, Independent Component Analysis (ICA) which will directly or/and indirectly used to recognize human expression in various condition.

Index Terms: Facial Expression, Expression Recognition, Gabor Filter, Gabor Energy, Principal Component Analysis, Neural Network, Eigenface.

Abstract: In this paper, the exploration of face recognition technology that is 2D face recognition is being analyzed. Face Recognition is widely used for security at many places like airport, organizations, many devices etc. The various implementation approaches widely accepted is been discussed. Each process in the face recognition consists of sub-process and the sub-process is categorized into registration, representation, extraction of discriminative features.

Index Terms: Face recognition, Eigen-face, Fisher-face, Euclidean distance

Abstract: The rapid growth of digital world and networked multimedia systems has generated an urgent need for copyright protection and ownership authentication. Digital watermarking techniques are one such technology that has been developed to protect digital media from an illegal manipulation. Particular, in this paper we introduce and simulated two algorithms of digital watermarking technique. The first technique is based on DCT (Discrete Cosine Transform and second technique is based on DWT (Discrete Wavelet Transform). The simulation results are shown and compared for different quality.

Index Terms: Image Processing, Watermarking, DCT (Discrete Cosine Transform), DWT (Discrete Wavelet Transform), Frequency Domain Watermarking.

Abstract: Recognizing text from scanned document is the classical problem of pattern recognition and image processing. In this paper, we have proposed hierarchical algorithm for English Optical Character Recognition (OCR). Beauty of this algorithm is that it does not need any features or classifiers like Neural Network (NN) or Support Vector Machine (SVM). Proposed approach recognizes various characters at different level of its hierarchy using Euler property, symmetry and histogram. We have exploited geometric shape of characters for recognition process.

Index Terms: Optical Character Recognition, Euler Number, Histogram, Vertical Symmetry, Horizontal Symmetry.

Abstract: Optical Character Recognition Systems are getting more and more attention in recent decade. In many countries, OCR has been a part of their government sectors like post offices, Library automation, License Plate Recognition, Defence organization etc. According to recent survey, there are at least 550 million people are using Devanagari script for communication. Hindi is one of the language, which is derived from Devanagari script. For any character recognition system, essential step is to identify individual character and find features to compare it with the template features. In this paper, we have proposed histogram based hierarchical approach for isolating individual character from the image document. We have used Principle Component Analysis and Fisher Discriminant Analysis kind of holistic features for recognition. We have done the comparisons of such holistic features with geometric features like binary features and chain code.

Index Terms: Pre-processing, Segmentation, Histogram, Neural Network, Support Vector Machine

Abstract: In this paper, the exploration of new face recognition technology that is 3D face recognition is being analyzed. Face Recognition is widely used for security at many places like airport, organizations, many devices etc. The challenges faced in 2D face recognition technology is been solved through various approaches mentioned in the paper. The various implementation approaches widely accepted is been discussed. Each process in the face recognition consists of sub-process and the sub-process is categorized into registration, representation, extraction of discriminative features.

Index Terms: Face recognition, Eigen-Surface, Fisher-surface, Euclidean distance

Abstract: Pattern recognition deals with categorization of input data into one of the given classes based on extraction of features. Handwritten Character Recognition (HCR) is one of the well-known applications of pattern recognition. For any recognition system, an important part is feature extraction. A proper feature extraction method can increase the recognition ratio. In this paper, a chain code based feature extraction method is investigated for developing HCR system. Chain code is working based on 4-neighborhood or 8–neighborhood methods. In this paper, 8–neighborhood method has been implemented which allows generation of eight different codes for each character. These codes have been used as features of the character image, which have been later on used for training and testing for Neural Network (NN) and Support Vector Machine (SVM) classifiers. In this work we have also implemented HCR system with the use of correlation coefficient. Comparison of all the methods for HCR systems are highlighted at the end.

Index Terms: Pattern recognition, handwritten character recognition, feature extraction, chain code, correlation coefficient, neural network, support vector machine.

Abstract: Facial Expression Recognition (FER) is a rapidly growing and ever green research field in the area of Computer Vision, Artificial Intelligent and Automation. There are many application which uses Facial Expression to evaluate human nature, feelings, judgment, opinion. Recognizing Human Facial Expression is not a simple task because of some circumstances due to illumination, facial occlusions, face color/shape etc. In these paper, we present some method/techniques such as Principal Component Analysis (PCA), Linear Discriminate Analysis (LDA), Gabor Filter/Energy, Line Edge Mapping (LEM), Neural Network, Independent Component Analysis (ICA) which will directly or/and indirectly used to recognize human expression in various condition.

Index Terms: Facial Expression, Expression Recognition, Gabor Filter, Gabor Energy, Principal Component Analysis, Neural Network, Eigenface.

Abstract: In the world of digital media, digital watermark technique is widely applied to tampering detection, authenticity and/or ownership protection of digital images, audio, video or even texts. Digital watermarking received increasing attention in the last decade due to massive digital artwork distribution via internet. It is the process of embedding information into digital signals. Watermarking is very common in our everyday lives, we see watermarking in currency, government documents, stamps, television broadcast and many other common documents. Two main categories of watermarks are visible and invisible (imperceptible). Visible methods provide means for overt assertion of ownership with logos; the invisible methods provide covert protection of these rights. There are two main domains for digital watermarking technique: spatial domain and Frequency / Transformed Domain. In this paper a basic concept of digital watermarking, properties and requirements, various classifications and existing techniques such as LSB, DCT, DFT, DWT, SVD and many more are reviewed.

Index Terms: Watermarking, Steganography, Properties of Digital Image Watermarking, Spatial Domain, Existing Watermarking techniques.

Abstract: According to recent survey, there are at least 550 million people are using Devnagari script for communication. Hindi is one of the language, which is derived from Devnagari script. As it is a national language of India, Hindi Optical Character Recognition (OCR) System has a wide application in areas like post offices, Library automation, License Plate Recognition, Defense organization and many more government sectors. Research on printed and handwritten character recognition has been started before 50 years. Most of the research has been done for European texts. For any character recognition system, essential step is to identify individual character and find features to compare it with the template features. In this paper, we have proposed histogram based hierarchical approach for isolating individual character from the image document. For recognizing the characters, we can use Euclidean distanc, Pearson coefficient, chain code etc. Result shows that our system is quite robust and provides accuracy up to 92% for the character isolation.

Index Terms: Preprocessing, Segmentation, Skew correction, Histogram, Shirorekha.

Abstract: In this paper, we have discussed dimensionality reduction techniques for face recognition – Principle Component Analysis (PCA) and Fisher Discriminant Analysis (FDA). Both the methods are based on linear projection, which projects the face from higher dimensional image space to lower dimensional feature space. PCA derives the most expressive features (MEF) by projecting face vector such that it captures greatest variance. FDA derives most discriminating features (MDF) by maximizing between class scatter and minimizing within class scatter. Lower dimensional features are used for recognition process. Classification can be achieved using Neural Network (NN), Support Vector Machine (SVM) etc. We have tested our system for the L2 norm measure. At the end of the paper, we have discussed results which show that FDA out weights the performance of PCA with average recognition rate more than 95 %.

Index Terms: Face recognition, Eigen face, Eigen Vector, Scatter matrix, Fisher Face.

Abstract: In this paper, we propose an algorithm to detect semantic concepts from cricket video. In our previous work, we have proposed key frame detection based approach for semantic event detection and classification. The proposed scheme works in two parts. In first part a top-down event detection and classification is performed using hierarchical tree. In second part, higher level concept is identified by applying A-Priori algorithm. In part 1, key frames are identified based on Hue Histogram difference at level 1. At level 2, logo transitions classify the frames as real-time or replay. At level 3, we classify the real time frames as field view, pitch view or non field view based on thresholds like Dominant Soli Pixel Ration (DSPR) and Dominant Grass Pixel Ration (DGPR). At level 4, we detect close up and crowd frames based upon edge detection. At level 5a, we classify the close up frames into player of team A, player of team B and umpire based upon skin colour and corresponding jersey colour. At level 5b, we classify the crowd frames into spectators, player’s gathering of team A or player’s gathering of team B. In part two, labels are associated with each frame event, which is used as input to A-Priori algorithm for concept mining. Results at the end of paper show the robustness of our approach.

Index Terms: Histogram, Dominant Grass Pixel Ratio, Dominant Soil Pixel Ratio, Concept Mining, A-Priori Algorithm

Abstract: Character recognition plays an important role in many recent applications. It can solve more complex problem and make human job easier. An example is Handwritten Character Recognition (HCR). This system is widely used to interpret zip codes, postal addresses, bank checks, and many other applications. There are different techniques are available for handwritten character recognition. In this paper, a review of different techniques used for handwritten character recognition systems is presented. Different techniques used for different phases of HCR such as segmentation, feature extraction, and classification are reviewed and analyzed in this work. Some of the challenges available in using HCR systems in real-time applications are highlighted at the end.

Index Terms: Handwritten character recognition, segmentation, feature extraction, classification.

Abstract: Character recognition plays an important role in many recent applications. It can solve more complex problem and make human job easier. An example is Handwritten Character Recognition (HCR). This system is widely used to interpret zip codes, postal addresses, bank checks, and many other applications. There are different techniques are available for handwritten character recognition. In this paper, a review of different techniques used for handwritten character recognition systems is presented. Different techniques used for different phases of HCR such as segmentation, feature extraction, and classification are reviewed and analyzed in this work. Some of the challenges available in using HCR systems in real-time applications are highlighted at the end.

Index Terms: Handwritten character recognition, segmentation, feature extraction, classification

Abstract: Nowadays image authentication plays a vital role for security. Watermarking techniques facilitate to hide the data in carrier image in such a way so that data would become imperceptible. Watermarking can be performed either in spatial domain or in transfer domain. Least Significant Bit (LSB) modification is the simple but non reliable theme, which is performed in spatial domain. In transfer domain, Singular Value Decomposition (SVD), Discrete Cosine Transformation (DCT), Fast Fourier Transformation (FFT), Discrete Wavelet Transformation (DWT) and Fast hadamard Transform (FHT) etc provide robust result. The goal of the paper is to present how such methods are useful to authenticate images and avoid image tampering. In our work, we have discussed block and non block based watermarking scheme performed in DCT domain using SVD. Several experiments are carried out to check the robustness of the algorithm. We conclude that non block based algorithm is superior in terms of visual quality after embedding and extraction, while its inferior in terms of some common signal processing attacks like rotation, scaling, translation etc.

Index Terms: Watermarking, Transform domain, Discrete Cosine Transform, Singular Value Decomposition, Singular Values

Abstract: In this paper, we propose a key frame detection based approach towards semantic event detection and classification in cricket videos. The proposed scheme performs a top-down event detection and classification using hierarchical tree. At level 1, we extract key frames for indexing based upon the Hue Histogram difference. At level 2, we detect logo transitions and classify the frames as realtime or replay fragments. At level 3, we classify the realtime frames as field view, pitch view or non field view based on colour features such as soil colour for pitch view and grass colour for field view. At level 4, we detect close up and crowd frames based upon edge detection. At level 5a, we classify the close up frames into player of team A, player of team B and umpire based upon skin colour and corresponding jersey colour. At level 5b, we classify the crowd frames into spectators, player’s gathering of team A or player’s gathering of team B. Our classifiers show excellent results with correct detection and classification with reduced processing time.

Index Terms: Histogram, Template matching, Dominant Grass Pixel Ratio, Dominant Soil Pixel Ratio, Connected Component Analysis

Abstract: Image compression is a method through which we can reduce the storage space of images, videos which will helpful to increase storage and transmission process’s performance. In image compression, we do not only concentrate on reducing size but also concentrate on doing it without losing quality and information of image. In this paper, two image compression techniques are simulated. The first technique is based on Discrete Cosine Transform (DCT) and the second one is based on Discrete Wavelet Transform (DWT). The results of simulation are shown and compared different quality parameters of its by applying on various images.

Index Terms: DCT, DWT, Image compression, Image processing

Abstract: Since last decade, face recognition has replaced almost all biometric authentication techniques available. Many algorithms are in existence today based on various features. In this paper, we have compared the performance of various classifiers like correlation, Artificial Neural Network (ANN) and Support Vector Machine (SVM) for Face Recognition. We have proposed face recognition based on discriminative features. Holistic features based methods Fisher Discriminant Analysis (FDA) is used to extract out discriminative features from the input face image respectively. These features are used to train classifiers like Artificial Neural Network (ANN) and Support Vector Machine (SVM). Results in the last section describe the accuracy of proposed scheme.

Index Terms: Face Recognition, Fisher Discriminant

Abstract: Pattern recognition deals with categorization of input data into one of the given classes based on extraction of features. Handwritten Character Recognition (HCR) is one of the well-known applications of pattern recognition. For any recognition system, an important part is feature extraction. A proper feature extraction method can increase the recognition ratio. In this paper, a Principal Component Analysis (PCA) based feature extraction method is investigated for developing HCR system. PCA is a useful statistical technique that has found application in fields such as face recognition and image compression, and is a common technique for finding patterns in data of high dimension. These method have been used as features of the character image, which have been later on used for training and testing with Neural Network (NN) and Support Vector Machine (SVM) classifiers. HCR is also implemented with PCA and Euclidean distance.

Index Terms: Pattern recognition, handwritten character recognition, feature extraction, principal component analysis, neural network, support vector machine, euclidean distance.

Abstract: Many multimedia applications and entertainment industry products like games, cartoons and film dubbing require speech driven face animation and audio-video synchronization. Only Automatic Speech Recognition system (ASR) does not give good results in noisy environment. Audio Visual Speech Recognition system plays vital role in such harsh environment as it uses both – audio and visual – information. In this paper, we have proposed a novel approach with enhanced performance over traditional methods that have been reported so far. Our algorithm works on the bases of acoustic and visual parameters to achieve better results. We have tested our system for English language using LPC, MFCC and PLP parameters of the speech. Lip parameters like lip width, lip height etc are extracted from the video and these both acoustic and visual parameters are used to train systems like Artificial Neural Network (ANN), Vector Quantization (VQ), Dynamic Time Warping (DTW), Support Vector Machine (SVM). We have employed neural network in our research work with LPC, MFCC and PLP parameters. Results show that our system is giving very good response against tested vowels.

Index Terms: Automatic Speech Recognition, Phoneme, Viseme, Speech Parameter, Neural Network.

Abstract: Speech processing has vast application in voice dialing, telephone communication, call routing, domestic appliances control, Speech to text conversion, text to speech conversion, lip synchronization, automation systems etc [1]. Nowadays, Speech processing has been evolved as novel approach of security. Feature templates of authorized users are stored in database. Speech features are extracted from test speech and compared with templates available in database. Speech can be parameterized by Linear Predictive Codes (LPC), Perceptual Linear Prediction (PLP), Mel Frequency Cepstral Coefficients (MFCC) PLP-RASTA (PLP-Relative Spectra) etc. Some parameters like PLP and MFCC considers the nature of speech while it extracts the features, while LPC predicts the future features based on previous features. We have practically analyzed this truth in our research work. Training models like neural network are trained for feature vector to predict the unknown sample. Techniques like Vector Quantization (VQ), Dynamic Time Warping (DTW), Support Vector Machine (SVM) can be used for template matching and classification. We have employed neural network in our research work with LPC, PLP and MFCC parameters.

Index Terms:

Abstract: Face recognition has been grown as a prime security idea since last decade. Face detection is the basic step in face recognition. In this paper, we have discussed the approach to detect faces from the 2D color images with single or multiple faces. Proposed algorithm work in three steps, skin color segmentation, morphological operations and last step is face rejection or acceptance. The beauty of this algorithm is that it is scale independent and orientation invariant. Skin color segmentation works as a preprocessing step to reduce the processing space. Accuracy of this algorithm is checked against various images with dynamic condition, which shows accuracy range of 90% to 100 %.

Index Terms: Face detection, Hue, Saturation, Connected Component Analysis, Morphological operations.

Abstract: Many multimedia applications and entertainment industry products like games, cartoons and film dubbing require speech driven face animation and audio-video synchronization. Only Automatic Speech Recognition system (ASR) does not give good results in noisy environment. Audio Visual Speech Recognition system plays vital role in such harsh environment as it uses both – audio and visual – information. In this paper, we have proposed a novel approach with enhanced performance over traditional methods that have been reported so far. Our algorithm works on the bases of acoustic and visual parameters to achieve better results. We have tested our system for English language using MFCC and LPC parameters of the speech. Lip parameters like lip width, lip height etc are extracted from the video and these both acoustic and visual parameters are used to train neural network. Our system is giving almost cent percent response against vowels.

Index Terms: ASR, Phoneme, Viseme, Speech parameters, Neural Network.

Abstract: Speech processing has vast application in voice dialing, telephone communication, call routing, domestic appliances control, Speech to text conversion, text to speech conversion, lip synchronization, automation systems etc [1]. Nowadays, Speech processing has been evolved as novel approach of security. Feature templates of authorized users are stored in database. Speech features are extracted from test speech and compared with templates available in database. Speech can be parameterized by Linear Predictive Codes (LPC), Perceptual Linear Prediction (PLP), Mel Frequency Cepstral Coefficients (MFCC) PLP-RASTA (PLP-Relative Spectra) etc. Some parameters like PLP and MFCC considers the nature of speech while it extracts the features, while LPC predicts the future features based on previous features. We have practically analyzed this truth in our research work. Training models like neural network are trained for feature vector to predict the unknown sample. Techniques like Vector Quantization (VQ), Dynamic Time Warping (DTW), Support Vector Machine (SVM) can be used for template matching and classification. We have employed neural network in our research work with LPC, PLP and MFCC parameters.

Index Terms:

Abstract: To identify the individual using physical features has got great attention since long. Today Ear Recognition has been introduced as one of the very good biometric technique for security [1]. At many places, ear recognition has played vital role in forensic science too. Lot many reasons and strong background is available for selecting ear for security purpose. Ear recognition has out weighted all traditional biometric systems like Face recognition, finger biometric, Iris biometric etc. Ear recognition is non invasive biometric and hence sometimes it performs better than iris or finger biometric in noisy environment. We have employed Force Field Feature Extraction Algorithm in our research. We have extracted ear features using proposed algorithm. In this research work, we are directly comparing the features to identify weather person belongs to given database or not. Neural network can be trained for the classification and identification of individual. This algorithm provides robust results for small database and hence it has practical application in small firms and organizations.

Index Terms:

Abstract: In this paper, we have presented an approach for lip detection and lip feature extraction. Lip detection and tracking has great application in animation, movies, news reading, and film industries. Lots of work has been done in area of various face component detection and recognition. Apart from eye detection, ear detection, iris detection etc, lip tracking and lip detection is one of the favourite topics for researchers. Various algorithms and techniques have been implemented so for to achieve better and better performance. Normalized RGB color scheme, HSV color model, Lip detection using HUE segmentation and many more techniques have been implemented and are in the boom. All methods are having their own pros and cons. We are aiming to detect lips and extract the lip features like width and height from face using hue, luminance, and chrominance component of the image. The reason behind the selection of this algorithm is that, it performs well under various illumination conditions, which is the one of the dimension of difficulty in the area of lip detection. We have carried out the work on inhouse database with varying lighting and noisy conditions.

Index Terms: Hue, Eye localization, Lip reading, lip tracking, HSV Color Space.

Abstract: This paper represents the basic idea about lip synchronization and some techniques to how to map the real time voice on lip. Lip synchronization has very vast application in game and film industries. Before the emergence of various lip models, the easiest approach to display lip animation is to use prestored images to represent all the possible shapes and combine these with for example some morphing method to smooth boundaries. This method may be quite successful for some limited applications, but it is very inflexible, since there is no way to control different facial features independently of each other. Lip features such as lip width, inner lip height, and outer lip height are acquired by some extraction algorithms, that uses both color information and edges. LSP (Linear Spectrum Pair) / LPC (Linear Prediction Coefficient) / MFCC (Mel Scale Frequency Cepstral Coefficient) coefficients are used to parameterize the speech. LSP / LPC / MFCC coefficients and lip features are used to train the mapping model like Hidden Markov Model, Neural Network etc. Resulting models are used to estimate lip features from acoustic speech.

Index Terms:

Abstract: For the first time, thanks to the increased power of computers, new methods replace the skill of the statistical artisan with massive-computational methods, obtaining equal or better results in far less time without the need for any specialized knowledge. Data Mining is probably the most useful way to take advantage of the massive processing power available on computers, and the definitely most promising and exciting research field in Advanced Informatics. Neural Networks algorithms are among the most popular data mining and machine learning techniques used today. As computers become faster, the neural net methodology is replacing many traditional tools in the field of knowledge discovery and some related fields.

Index Terms: