&ldquo;IIT Jodhpur is committed to impart quality education to facilitate the creation of qualified, competent and responsible human resources to meet the emerging technological challenges of the world. It is working towards creating an equitable learning platform that is positive, accessible and effective in the face of the modern societal and scientific advancements.

Flexibility in curriculum, in all its academic programs, encompassing elements of teaching, research, creativity, innovation and entrepreneurship is the most unique feature of IIT Jodhpur. The Department of Science and Technology (DST), under the National Mission on Cyber Physical Systems, has identified IIT Jodhpur as a Technology &amp; Innovation Hub in the area of Augmented Reality and Virtual Reality. The institute has also been nominated as the coordinator for the Jodhpur City and Knowledge Cluster by the Office of Principal Scientific Adviser, Government of India.

The sprawling 852-acre IIT Jodhpur campus is a majestic tribute to the deep-rooted diverse cultural heritage of the region. The campus is an architectural masterpiece, which is slated to become an international exemplar of sustainability with net-zero energy, and advanced water and waste management strategies.&rdquo;

IIT Jodhpur

Biometrics is the process of verifying or identifying an individual using the person's physiological and behavioral characteristics such as face, fingerprint, iris, signature, and gait. The main advantage of using biometrics for recognition is that it is unique for every individual and cannot be misplaced or forgotten. However, biometric systems that use a single biometric trait have to contend with noisy data, restricted degrees of freedom, failure to-enroll problems, spoof attacks, and unacceptable error rates. Various researchers have suggested that no single biometric modality can provide the protection required for high security applications [1] - [4]. To alleviate this problem and enhance the performance of a biometric system, information from different biometric sources are combined and such systems are known as multimodal biometric systems [2]. © Springer-Verlag Berlin Heidelberg 2007.

Department of Computer Science & Engineering

World University root node

Studies in Fuzziness and Soft Computing

Intelligent biometric information fusion using support vector machine

This paper proposes an intelligent 2v-support vector machine based match score fusion algorithm to improve the performance of face and iris recognition by integrating the quality of images. The proposed algorithm applies redundant discrete wavelet transform to evaluate the underlying linear and non-linear features present in the image. A composite quality score is computed to determine the extent of smoothness, sharpness, noise, and other pertinent features present in each subband of the image. The match score and the corresponding quality score of an image are fused using 2v-support vector machine to improve the verification performance. The proposed algorithm is experimentally validated using the FERET face database and the CASIA iris database. The verification performance and statistical evaluation show that the proposed algorithm outperforms existing fusion algorithms. © World Scientific Publishing Company.

International Journal of Neural Systems

Integrating image quality in 2v-ASVM biometric match score fusion

In this paper, we propose a 2-stage preprocessing framework which consists of image enhancement and deformation techniques to increase the verification performance of image-based biometric systems. In the preprocessing framework, first the quality of biometric image is enhanced and then a deformation model is applied to minimize the variation between the two images to be matched. The proposed SVM image quality enhancement algorithm selects good quality regions from different globally enhanced images and combines them to generate a single high-quality feature-rich image. The proposed deformation algorithm is based on phase congruency information and locally minimizes the variations between two images while retaining the features required for recognition. The proposed algorithms are validated using face and iris biometrics as the two case studies. For performance evaluation, different face and iris recognition algorithms are chosen and the verification accuracy is computed using non-homogenous face and iris databases. Experimental results show that the performance of face and iris recognition algorithms are significantly improved when the input images are preprocessed using the proposed enhancement and deformation algorithms. © 2007 Elsevier B.V. All rights reserved.

Signal Processing

Improving verification accuracy by synthesis of locally enhanced biometric images and deformable model

This paper presents a novel age transformation algorithm to handle the challenge of facial aging in face recognition. The proposed algorithm registers the gallery and probe face images in polar coordinate domain and minimizes the variations in facial features caused due to aging. The efficacy of the proposed age transformation algorithm is validated using 2D log polar Gabor based face recognition algorithm on a face database that comprises of face images with large age progression. Experimental results show that the proposed algorithm significantly improves the verification and identification performance. © Springer-Verlag Berlin Heidelberg 2007.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Age transformation for improving face recognition performance

This paper formulates an evidence theoretic multimodal fusion approach using belief functions that takes into account the variability in image characteristics. When processing non-ideal images the variation in the quality of features at different levels of abstraction may cause individual classifiers to generate conflicting genuine-impostor decisions. Existing fusion approaches are non-adaptive and do not always guarantee optimum performance improvements. We propose a contextual unification framework to dynamically select the most appropriate evidence theoretic fusion algorithm for a given scenario. The effectiveness of our approach is experimentally validated by fusing match scores from level-2 and level-3 fingerprint features. Compared to existing fusion algorithms, the proposed approach is computationally efficient, and the verification accuracy is not compromised even when conflicting decisions are encountered. ©2007 IEEE.

IEEE Conference on Biometrics: Theory, Applications and Systems, BTAS'07

Unification of evidence theoretic fusion algorithms: A case study in level-2 and level-3 fingerprint features

Mosaicing entails the consolidation of information represented by multiple images through the application of a registration and blending procedure. We describe a face mosaicing scheme that generates a composite face image during enrollment based on the evidence provided by frontal and semiprofile face images of an individual. Face mosaicing obviates the need to store multiple face templates representing multiple poses of a user's face image. In the proposed scheme, the side profile images are aligned with the frontal image using a hierarchical registration algorithm that exploits neighborhood properties to determine the transformation relating the two images. Multiresolution splining is then used to blend the side profiles with the frontal image, thereby generating a composite face image of the user. A texture-based face recognition technique that is a slightly modified version of the C2 algorithm proposed by Serre et al. is used to compare a probe face image with the gallery face mosaic. Experiments conducted on three different databases indicate that face mosaicing, as described in this paper, offers significant benefits by accounting for the pose variations that are commonly observed in face images. © 2007 IEEE.

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

A mosaicing scheme for pose-invariant face recognition

This paper presents a novel digital watermarking technique using face and demographic text data as multiple watermarks for verifying the chain of custody and protecting the integrity of a fingerprint image. The watermarks are embedded in selected texture regions of a fingerprint image using discrete wavelet transform. Experimental results show that modifications in these locations are visually imperceptible and maintain the minutiae details. The integrity of the fingerprint image is verified through the high matching scores obtained from an automatic fingerprint identification system. There is also a high degree of visual correlation between the embedded images, and the extracted images from the watermarked fingerprint. The degree of similarity is computed using pixel-based metrics and human visual system metrics. The results also show that the proposed watermarked fingerprint and the extracted images are resilient to common attacks such as compression, filtering, and noise. © 2006 Elsevier Ireland Ltd. All rights reserved.

Forensic Science International

Enhancing security of fingerprints through contextual biometric watermarking

This paper presents a novel fingerprint classifier fusion algorithm using Dempster-Shafer theory concomitant with update rule. The proposed algorithm accurately matches fingerprint evidences arid also efficiently adapts to dynamically evolving database size without compromising accuracy or speed. We experimentally validate our approach using three fingerprint recognition algorithms based on minutiae, ridges, and image pattern features. The performance of our proposed algorithm is compared with these individual fingerprint algorithms and commonly used fusion algorithms. In all cases, the proposed Dempster Shafer theory with update rule outperforms existing algorithms even with partial fingerprint image. We also show that as the database size increases, the proposed algorithm is designed to operate on only the augmented data instead of the entire database, thereby reducing the training time without compromising the verification accuracy. © IEICE 2006.

IEICE Electronics Express

DS theory based fingerprint classifier fusion with update rule to minimize training time

This paper presents a combined DWT and LSB based biometric watermarking algorithm that securely embeds a face template in a fingerprint image. The proposed algorithm is robust to geometric and frequency attacks and protects the integrity of both the face template and the fingerprint image. Experimental results performed on a database of 750 face and 750 fingerprint images show that the algorithm has the advantages of both the existing DWT and LSB based algorithms. A multimodal biometric algorithm is used as a metric to evaluate the combined performance of both face and fingerprint recognition. © IEICE 2006.

Robust biometric image watermarking for fingerprint and face template protection

A novel face recognition algorithm using single training face image is proposed. The algorithm is based on textural features extracted using the 2D log Gabor wavelet. These features are encoded into a binary pattern to form a face template which is used for matching. Experimental results show that on the colour FERET database the accuracy of the proposed algorithm is higher than the local feature analysis (LFA) and correlation filter (CF) based face recognition algorithms even when the number of training images is reduced to one. In comparison with recent single training image based face recognition algorithms, the proposed 2D log Gabor wavelet based algorithm shows an improvement of more than 3% in accuracy. © IEE 2005.

Electronics Letters

Textural feature based face recognition for single training images

This paper presents a novel biometric watermarking algorithm for improving the recognition accuracy and protecting the face and fingerprint images from tampering. Multi-resolution Discrete Wavelet Transform is used for embedding the face image in a finger print image. An intelligent learning algorithm based on õ-Support Vector Machine (SVM) is introduced to enhance the quality of the extracted face image. The performance of the watermarking algorithm is experimentally validated using existing fingerprint and face recognition algorithms. The results show that the extracted fingerprint and faceimages are of high quality. The use of SVMenhances the performance of face recognition by at least 10% even when the watermarked image is subjected to certain geometric and frequency attacks such as scaling, cropping, compression and filtering. © 2005, The Institute of Electronics, Information and Communication Engineers. All rights reserved.

ieice electronics express

Improving biometric recognition accuracy and robustness using DWT and SVM watermarking

We describe a face mosaicing scheme that generates a composite face image during enrollment based on the evidence provided by frontal and semi-profile face images of an individual. Face mosaicing obviates the need to store multiple face templates representing multiple poses of a user's face image. In the proposed scheme, the side profile images are aligned with the frontal image using a terrain transform that exploits neighborhood properties to determine the transformation relating the 2 images. Multiresolution splining is then used to blend the side profiles with the frontal image thereby generating a composite face image of the user. A local binary pattern matching algorithm is used to compare an input face image with the template face mosaic. Experiments conducted on a small dataset of 27 users indicate that face mosaicing as described in this paper offers significant benefits by accounting for the pose variations commonly observed in face images.

Proceedings - Fourth IEEE Workshop on Automatic Identification Advanced Technologies, AUTO ID 2005

Performance enhancement of 2D face recognition via mosaicing

A signature verification algorithm based on static and dynamic features of online signature data is presented. Texture and topological features are the static features of a signature image whereas the digital tablet captures in real-time the pressure values, breakpoints, and the time taken to create a signature. ID - log Gabor wavelet and Euler numbers are used to analyze the textural and topological features of the signature respectively. A multi-classifier decision algorithm combines the results obtained from three feature sets to attain an accuracy of 98.18%. © Springer-Verlag Berlin Heidelberg 2004.

Signature verification using static and dynamic features

In this paper, we have studied various well known algorithms for iris recognition. Four algorithms due to Avila [5], Li Ma [9], Tisse [13] and Daugman [15] are implemented and compared on the CASIA iris image database. The results show that the Daugman's algorithm gave the highest accuracy of 99.9%.

Proceedings of International Conference on Intelligent Sensing and Information Processing, ICISIP 2004

Comparison of Iris Recognition Algorithms

In this paper we have presented a fusion of three face recognizers, LFA, LBA and KDDA, -which combines the three Confidence Measure Factors (CMFs) using a RBF neural netwrok. This strategy is used to increase the accuracy of the face recognition system. The recognizer is tested on AT&amp;T-ORL, AR and IITK Face database and the results are found to be more than 95%. © 2004 IEEE.

Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

Face recognition using multiple recognizers

This paper presents a multimodal biometrics system using watermarking algorithms with two levels of security for simultaneously verifying an individual and protecting the biometric template. Iris template is watermarked in face, such that the face is visible for verification and the watermarked iris is used to cross authenticate the individual and secure the biometrics data as well. The accuracy of the multimodal biometrics system is around 96.8%. This system is also resistant to common attacks on biometric templates. © 2004 IEEE.

Digital watermarking based secure multimodal biometric system

In this paper three algorithms for iris verification have been presented. Iris detection algorithms include the normalization and iris extraction steps. Three algorithms for verification process are (a) Algorithm using radial and circular features, (b) Algorithm using Fourier transforms and (c) Algorithm using Circular-Mellin transforms. Proposed algorithms have been tested on CASIA database and some non-infrared Iris images. The experimental results show that the algorithm based on Circular - Mellin Transform gives the best result with an accuracy of 95.45%. Some initial experiments on non-infrared iris images shows that this algorithm can work on such images but it still requires some more attention and this is our future work. © Springer-Verlag Berlin Heidelberg 2004.

Iris based human verification algorithms

In this paper, a detailed experimental study of face detection algorithms based on "Skin Color" has been made. Three color spaces, RGB, YCbCr and HSI are of main concern. We have compared the algorithms based on these color spaces and have combined them to get a new skin color based face detection algorithm which gives higher accuracy. Experimental results show that the proposed algorithm is good enough to localize a human face in an image with an accuracy of 95.18%.

Tamkang Journal of Science and Engineering

A robust skin color based face detection algorithm

Face detection is an important research area having wide applications in man-machine interface, visual surveillance and face recognition. We have proposed an algorithm for face detection using Gradient Vector Flow in gray level images which overcomes the problem for localization and initialization. The algorithm has been tested on various face databases and the result shows an accuracy of 97% with invariance to pose and orientation.

International Conference on Machine Learning and Cybernetics

Face detection using gradient vector flow

One of the most successful applications of image analysis and understanding, face recognition has received significant attention. There are at least two reasons for the trend: the first is the wide range of commercial and law enforcement application and the second is the availability of feasible technologies. In general, few methods of face recognition are in practice, Feature based face recognition methods, Eigen Face Based, Line Based, Elastic Bunch Graph method and Neural Network based methods. All have their possibilities and features. In Neural Network approach automatic detection of eyes and mouth is followed by a spatial normalization of the images. The classification of the normalized images is carried out by hybrid Neural Network that combines unsupervised and supervised methods for finding structures and reducing classification errors respectively. The line-based is a type of image-based approach. It does not use any detailed biometric knowledge of the human face. These techniques use either the pixel-based bi-dimensional array representation of the entire face image or a set of transformed images or template sub-images of facial features as the image representation. An image-based metric such as correlation is then used to match the resulting image with the set of model images. In the context of image-based techniques, two approaches are there namely template-based and neural networks. In the template-based approach, the face is represented as a set of templates of the major facial features, which are then matched with the prototypical model face templates. Neural network-based image techniques use an input image representation that is the gray-level pixel-based image or transformed image which is used as an input to one of a variety of neural network architectures, including multilayer, radial basis functions and auto-associative networks.