International Journal of Biometrics (14 papers in press)

Regular Issues

• A Secure Finger vein Recognition System using WS-Progressive GAN and C4 Classifier  
  by Sreemol R., Santosh Kumar M. B, Sreekumar A 
  Abstract: This paper proposes a secure finger vein reconstruction and recognition system utilising a novel weight standardisation-based progressive generative adversarial networks (WS-progressive GAN) as well as ' he' initialised chimp optimisation-based convolutions neural network (he-ChOA-CNN) classifier for overcoming security issues. Initially, the input images are pre-processed, and the reflection-based contrast limited adaptive histograms equalisation (RCLAHE) enhanced the pre-processed images. Next, bias locality-sensitive hashing (BLSH) generates hash values, through which the ameliorated images are secured. Next, the secured images are augmented and applied for WS-progressive GAN, which encodes and decodes the image for reconstructing the synthetic images. Then, the he-ChOA-CNN accepts the imperative features extracted as of the synthetic images as input for training. Amid testing, the identity of the person is recognised utilising the classifier output and the query image by detecting the gaps. Analogised to the prevailing methods, more accurate outcomes are attained by the proposed model, which is illustrated through the experimental outcomes.
  Keywords: reflection-based contrast limited adaptive histogram equalisation; finger vein; bias locality-sensitive hashing; BLSH; he initialisation; chimp optimisation-based CNN; generative adversarial network.
  DOI: 10.1504/IJBM.2024.10053719
   
  
• Recent trends and challenges in human computer interaction using automatic emotion recognition: a review  
  by Sukhpreet Kaur, Nilima Kulkarni 
  Abstract: Automatic emotion recognition (AER) using facial expressions and electroencephalogram (EEG) signals is an interesting and booming area of research in the field of human computer interaction. This paper aims to identify the key state-of-art methodologies, understanding the standard workflow pipeline and knowing the existing findings. Different machine learning & deep learning approaches used recently for information preprocessing, feature extraction, feature classification and fusion schemes have also been explored. Furthermore, the purpose of this review work is to discuss the aspects motivating researchers to move from unimodal to multimodal AER systems. Also, this surveyed information is summarized in tabular forms to investigate the recent methods used and the results obtained. This comprehensive literature survey identifies the key points for inclusion of facial expressions and EEG signals over other channels. Also, the benefits of automated features which are being leveraged over hand crafted features for building improved real time emotion recognition systems.
  Keywords: Emotion recognition; Human computer interaction; Affective computing; Facial expressions;  EEG signals; Multimodal system.
  DOI: 10.1504/IJBM.2024.10053960
   
  
• Arabic Offline writer identification on a new version of AHTID/MW database  
  by Anis Mezghani Mezghani, Mongi Kherallah 
  Abstract: Handwriting is considered to be one of the commonly used biometric modality to verify and identify persons in commercial, governmental and forensic applications. In order to test and compare the accuracy of a computer vision system, in general, and a biometric system in particular, standard rich databases must be publicly available. In this paper and for this purpose, we expose the different works of writer identification of Arabic handwritten text carried out on our already published database AHTID/MW. As researchers have achieved high identification rates, we propose to extend the AHTID/MW database with new Arabic native writers and raise the level of difficulty. A baseline is drawn on each text-line image, and ground truth information is provided for each text image. In addition we present our experiments on the database using a new approach based on combining a CNN for feature extraction with GMM-based emission probability estimates for classification.
  Keywords: Arabic writer identification; handwritten text image; AHTID/MW database; convolutional neural network; Gaussian mixture model; GMM.
  DOI: 10.1504/IJBM.2024.10054549
   
  
• Exemplar-Based Facial Attribute Manipulation: A Review  
  by Padmashree G, Karunakar A.K 
  Abstract: Facial attribute manipulation gained a lot of attention when deep learning algorithms made amazing achievements during the last few years. Facial attribute manipulation is the process of combining or removing desired facial characteristics for a given image. Recently, generative adversarial networks (GANs) and encoder-decoder architecture have been used to tackle this problem, with promising results. We present a comprehensive overview of deep facial attribute analysis from the perspectives of manipulation using exemplars in this study. The model construction approaches, datasets, and performance evaluation measures that are frequently utilized are discussed. Following this, a review of various homogeneous and heterogeneous exemplar-based facial attribute manipulation algorithms is presented in detail. Furthermore, several other facial attribute-related issues and related applications in the real world, are also discussed. Lastly, we go over some of the issues that can arise as well as some interesting future research directions.
  Keywords: Facial attribute manipulation; Image generation; Deep Learning; Generative Adversarial Networks(GAN); Facial attributes; Generator; Discriminator.
  DOI: 10.1504/IJBM.2024.10054948
   
  
• Latent fingerprint segmentation using multi scale attention U-net  
  by AKHILA P, Shashidhar G. Koolagudi 
  Abstract: Latent fingerprints are the fingerprints lifted from the crime scene surfaces. Segmentation of latent fingerprints from the background is an important preprocessing task which is challenging due to the poor quality of the fingerprints. Though fingerprint segmentation approaches based on their orientation and frequency are reported in the literature, they could not adequately address the problem. We propose a latent fingerprint segmentation model based on the U-Net attention network in this work. We added the Atrous Spatial Pyramid Pooling (ASPP) layer to the network to facilitate multi-scale fingerprint segmentation. Our approach could effectively segment the latent fingerprint region from the background and even detect occluded and partial fingerprints with simple network architecture. To evaluate the performance, we have compared our results with the manual ground truth using NIST SD27A dataset. Our segmentation model has improved matching accuracy on the NIST SD27A dataset.
  Keywords: latent fingerprint segmentation; U-Net; attention; weighted cross entropy; multi-scale.
  DOI: 10.1504/IJBM.2024.10056003
   
  
• A Unique Approach Towards Keystroke Dynamics Based Entry-point User Access Control  
  by Soumen Roy, Devadatta Sinha, Rajat Pal, Utpal Roy 
  Abstract: Access control is an essential security service for computing devices, applications, and information. Among the different entry-point user access controls, keystroke dynamics (KDs) has gained popularity owing to its several merits, such as low cost, ease of usage, etc. In this study, we proposed a unique distance-based anomaly detector together with an appropriate template construction method leading to more realistic and accurate results. We validated our approach with ten standard datasets and compared the performance with 50 state-of-the-art anomaly detectors. In our consideration, recent anomaly detectors have been re-evaluated in the same experimental setting for sound comparison. An analysis of variance (ANOVA) was conducted to compare the performance of our approach to those detectors in both desktops and recent smartphones. This study provides an in-depth understanding of each detector’s performance which will aid in the design of efficient KD-based access control in the next generation of smart devices and applications.
  Keywords: access control; anomaly detection; keystroke dynamics; static authentication; template formation; template adaptation; touch dynamics; user authentication.
  DOI: 10.1504/IJBM.2024.10056899
   
  
• Fingerprint Multiple-Class Classifier: Performance Evaluation on Known and Unknown Fingerprint Spoofing Materials  
  by DIVINE AMETEFE, Suzi S. Sarnin, Darmawaty M. Ali, Dah John, Abdulmalik A. Aliu 
  Abstract: Fingerprint recognition is a popular and reliable biometric technology used in many security-sensitive applications. However, the use of fake fingerprints made from ubiquitous spoofing materials poses a significant threat to security systems. While several studies have proposed binary classifiers to detect fingerprint presentation attacks, relatively none have explored the effectiveness of multiple-class classifiers in detecting known and unknown spoofs. In this study, we evaluated the efficacy of multiple-class classifiers using deep transfer learning to detect presentation attacks made with different spoofing materials. Our experiments on the LivDet 2009 to 2015 datasets showed that while a classifier model developed without data augmentation performed better on known spoofs, it showed poor performance on cross-material detection of all seven fingerprint spoofing materials. These results suggest that modelling a multiple-class classifier is not an efficient approach for detecting cross-material presentation attacks in fingerprint recognition systems.
  Keywords: fingerprint spoofing; multiple-class classifier; known spoofing materials; unknown spoofing materials; deep transfer learning.
  DOI: 10.1504/IJBM.2024.10057333
   
  
• A comparative study on friction ridge pore features of males and females  
  by Anjana CD, Priyatha CV, Siva Prasad MS 
  Abstract: The sweat pores on the epidermal ridges of fingertips are unique and they are employed in personal identification. This study aimed to observe and analyse the pores within the left thumbprints of 50 individuals to find out whether there were any sex-related changes in the features of the pores. There was a significant (p < 0.05) difference in the average number of closed pores in males (53.60 ± 40.52) than that in females (81.60 ± 38.43). The average number of pores per 25 mm square was more frequent in females (156.12 ± 68.41) than males (105.12 ± 77.47). The difference in the distribution of pores per 5 mm of ridge length was found significant (p < 0.05) between males and females. The third-level features like type, shape, and frequency of the pores of males and females can be used as a presumptive indicator to determine the sex from fingerprints.
  Keywords: fingerprints; poroscopy; AFRS; personal identification; sex difference.
  DOI: 10.1504/IJBM.2024.10057334
   
  
• A Minutiae-Based Method to Store and Compare Fingerprints  
  by Eiman Alhamad, Mohammed Al Logmani, Abdullah Essa, Mohammad Hammoudeh 
  Abstract: Biometrics refers to certain physical or behavioural characteristics that are unique to every person. Biometrics, including fingerprints, are used for the measurement and analysis of biological data for identification purposes. This paper presents a new method to extract and compare fingerprint biometrics based on minutiae features. Only two reference minutiae are used to enhance the efficiency of the verification process with no need to match all the combinations of the extracted minutiae from the intellectual-reader with the reference minutiae in the alignment algorithm. The method is implemented and tested with an average decrease of 80% in the number of combinations required to be matched with the reference minutiae when two reference minutiae points are used instead of one to align and match fingerprints.
  Keywords: biometrics; authentication; fingerprint; minutiae.
  DOI: 10.1504/IJBM.2024.10059360
   
  
• A liveness detection system for sclera biometric applications  
  by Sumanta Das, Ishita De Ghosh, Abir Chattopadhyay 
  Abstract: Liveness detection systems are essential to test whether a biometric sample is from a live person. However, liveness detection for sclera biometric applications has not yet been investigated much. In a sensor-based approach, subjects are requested to view at specified directions. A gaze detection model LivGaze is proposed to verify whether the actual gaze direction matches with the requested one. A mismatch indicates an incorrect user response and hence a probable spoofing attack. In a feature-based approach, deep model LivDense is proposed for presentation attack detection. Three types of fake images are used for our work, namely, images scanned from printed papers, smart-phone display screens, and computer display screens. The two phases in a pipeline can be combined to form a system named LivSclera, which is efficient and cost-effective. We have achieved an average-case AUC of 0.987, accuracy of 0.99, and in the best-case 100% correct classifications on MASDUM dataset.
  Keywords: biometric; liveness; spoofing; sclera; PAD; LivGaze; LivDense; LivSclera; MASDUM; SBVPI.
  DOI: 10.1504/IJBM.2023.10050762
   
  
• PPG and fingerprint: robust bimodal biometric system  
  by Akhil Walia, Amit Kaul 
  Abstract: Technological advancements in the field of biometrics have resulted in the development of completely automatic methods for human recognition leading to a better and secure lifestyle. However, even though biometric traits like fingerprints exhibit a high degree of permanence, still certain security issues exist. The main objective of this work is to tackle spoofing attacks by introducing some liveness property in the biometric system. The PPG signal possesses various properties such as inherent liveness, ease of acquisition and low development cost. A robust biometric system, immune to direct attacks, using a combination of PPG and fingerprint has been suggested in this work. Score level fusion has been employed for integration of two modalities in parallel mode by optimising the scores of individual traits using interior point algorithm. This bimodal biometric with PPG and fingerprint as two modalities seems quite practical. A CRR of 100% has been achieved in experiments conducted on 38 healthy subjects.
  Keywords: fingerprint; photoplethysmogram; PPG; score level fusion; optimisation.
  DOI: 10.1504/IJBM.2023.10051187
   
  
• Use of synthetic signature images for biometric authentication and forensic investigation  
  by Sameera Khan, Megha Mishra, Vishnu Kumar Mishra 
  Abstract: Handwritten signatures are one of the widely used biometric traits for authentication, and are constantly questioned as forgery for this behavioural biometric is very common. Also due to privacy concerns, biometric databases are not easily available for training purposes. Due to this the efficiency of automatic authentication systems is highly compromised. The use of biometric data in forensic investigation also suffers from the problem of inadequate data. One of the solutions to this problem is the use of synthetic datasets in place of real datasets. Such datasets suffer from a high risk of generating unrealistic specimens. Generating high-quality synthetic biometric images is still a challenge. This paper discusses some of the basic requirements for synthetic signature generation and also proposes an algorithm to generate synthetic images for handwritten signatures using sinusoidal transformation.
  Keywords: synthetic signature; synthetic biometric; synthetic databases; forensic investigation.
  DOI: 10.1504/IJBM.2023.10050915
   
  
• Iris recognition system using deep learning techniques  
  by Amer A. Sallam, Hadeel Al Amery, Ahmed Y.A. Saeed 
  Abstract: Deep learning has been used and demonstrated intensively as a vital technique in data mining that can accurately and effectively evaluate enormous amounts of data for various applications. Iris recognition is one of those applications that necessitate complex algorithms for analysing and perfectly detecting the hidden patterns among its data in order to effectively distinguish one person from another. In this paper, an iris recognition system based on various deep learning techniques has been proposed. Through many experiments that were conducted on CASIA-V1 and ATVS datasets, the proposed system based on the Xception model was able to achieve significant results with 99.9% accuracy on CASIA-V1 dataset.
  Keywords: biometrics; deep learning; transfer learning; segmentation.
  DOI: 10.1504/IJBM.2023.10051537
   
  
• Deep learning with spectrogram image of eye movement for biometrics  
  by Antonio Ricardo Alexandre Brasil, Patrick Marques Ciarelli, Izabella Martins Da Costa Rodrigues, Jefferson Oliveira Andrade, Karin Satie Komati 
  Abstract: Biometric studies are being used worldwide for a large variety of purposes, here we used eye movements (EM) from observers of natural images. Since some EM are involuntary, these prevent spoofing attacks. While prior research requires feature extraction manually from EM data to identify a person, we use a deep convolutional architecture that processes it as an image. The eye movements were treated as a signal, then transformed as a spectrogram of frequencies, and its image is the input for a convolutional architecture. We investigated two types of signals: Cartesian coordinates, and gaze angle over time. The proposal consists of a convolutional network architecture applied to the DOVES dataset, where stimuli are natural images. We obtained the accuracy for the eye angle spectrogram, on DOVES, about 73%, and for the eye coordinates spectrogram, 65%. These results indicated that EM can be treated as spectrogram images for biometric identification.
  Keywords: DOVES dataset; eye angle; natural image stimuli.
  DOI: 10.1504/IJBM.2023.10052380