A Framework for Iris Partial Recognition based on Legendre Wavelet Filter

Danlami, Muktar; Jamel, Sapiee; Ramli, Sofia Najwa; Deris, Mustafa Mat

doi:10.14569/ijacsa.2019.0100506

Cited by 5 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The dataset comprises segmentation masks indicating the iris and pupil region boundaries, as well as metadata such as subject ID, gender, and age. Images are only partially captured because the subject's eyes are partially closed, the subject is looking away, or the subject's eyelashes partly covered the image of the subject's eyes [26].…”

Section: Mmu Iris Datasetmentioning

confidence: 99%

Online Multimodal Compression using Pruning and Knowledge Distillation for Iris Recognition

Samihah Abdul Latif,

Khairul Azami Sidek,

Eddyzulham Abu Bakar

et al. 2024

ARASET

View full text Add to dashboard Cite

Deep learning models have advanced to the forefront of image recognition tasks, resulting in high-performing but enormous neural networks with millions to billions of parameters. Yet, deploying these models in production systems imposes considerable memory limits. Hence, the research community is increasingly aware of the need for compression strategies that can reduce the number of model parameters and their resource requirement. Current compression techniques for deep learning models have limitations in efficiency and effectiveness, indicating that more research is required to develop more efficient and practical techniques capable of balancing the trade-offs between compression rate, computational cost, and accuracy. This study proposed a multimodal method by combining multimodal Pruning and Knowledge Distillation techniques for compressing the iris recognition model, which is the size constraint for many image recognition models. To maintain accuracy while shrinking the model’s size, the models are trained, compressed, and further retrained in the downstream job. The analysis includes both fully connected and convolutional layers. Experimentally, the findings show that the proposed technique can achieve 91% accuracy, the same as the existing or original model. Besides that, the model compression can reduce the size of the model almost six times, from 529MB to 90MB, which is a significantly reduced rate. The primary outcome of this study is developing a CNN lightweight model for iris recognition technology that can be used on mobile devices and is resource constrained.

show abstract

Section: Mmu Iris Datasetmentioning

confidence: 99%

Online Multimodal Compression using Pruning and Knowledge Distillation for Iris Recognition

Samihah Abdul Latif,

Khairul Azami Sidek,

Eddyzulham Abu Bakar

et al. 2024

ARASET

View full text Add to dashboard Cite

show abstract

“…The proposed system achieved 100% accuracy in a controlled environment; however, its efficiency in an uncontrolled environment is significantly lower (60%). In 2019, Danlami et al proposed a partial iris recognition system using a Legendre wavelet filter for partial iris feature extraction [12]. In 2020, Fu Liu et al proposed a partially occluded dorsal hand vein recognition system as an improved biometric graph matching method, including edge properties [13].…”

Section: Related Workmentioning

confidence: 99%

Effect of Smaller Fingerprint Sensors on the Security of Fingerprint Authentication

Jeong,

Jeong

2023

IEEE Access

View full text Add to dashboard Cite

Small fingerprint sensors are necessary to be used in the small devices like smartphones, USBs, and cards. Different small devices have different sizes of fingerprint sensors. The scanned fingerprint image through a small fingerprint sensor is a partial fingerprint image instead of a whole fingerprint image, if we use a small fingerprint sensor. If the size of the partial fingerprints stored in the database decreases, the number of minutiae that are compared when matching decreases. Unfortunately, for the small number of minutiae, it is possible to construct "Masterprints". Masterprints can be used to impersonate multiple users stored in the database. If the size of a partial image is 15% of a whole image, a Masterprint can be produced. We investigate the vulnerability of various sizes of partial images due to the Masterprint. Because sensors have different sizes, we use partial images of 9% to 48% of a whole image. Our results show that the security of partial fingerprint authentication is severely affected by the sizes of partial images and sensors. The smaller the partial image size, the larger the impostor matching rate; and the smaller the false matching rates, the steeper the impostor matching rate increases.

show abstract

“…In biometrics, iris became a highly informative object and high reliability as shown in Table 1, because the eye's highly distinctive qualities and easily accessible that do not alter throughout time [3,4]. Besides that, iris is not affected by age and does not require human contact with its scanning device [5]. Thus, this contactless biometric recognition gets attention from the public because they are relatively well known for their user-friendly and seamlessness [6].…”

Section: Introductionmentioning

confidence: 99%

An Efficient Iris Recognition Technique using CNN and Vision Transformer

Samihah Abdul Latif,

Khairul Azami Sidek,

Aisha Hassan Abdalla Hashim

2023

ARASET

View full text Add to dashboard Cite

The usage of biometric identification has increased in recent years, with numerous public and commercial organizations incorporating biometric technologies into their infrastructures. One of the technologies is iris recognition which has been used as a biometric recognition compared to other modalities to combat identity abuse due to its ability to eliminate risk of collisions or false matches even when comparing large populations. The use of CNN is proven to provide high accuracy; however, this technology involves the need for a large dataset and higher computational cost. Therefore, this study uses a combined model of Convolutional Neural Network (CNN) and Vision Transformer (ViT) in identifying and verifying an iris image. By using the proposed learning rate, it proves that the novel hybrid model is capable to achieve up to 93.66% accuracy in recognizing iris images. The cross-entropy loss function was implemented to reduce the loss and it was able to predict the class label more correctly. In addition, the model was thoroughly tested on three publicly available iris databases, achieving satisfactory iris recognition results. Furthermore, this model has the potential to be used in other biometrics such as face and retina recognitions.

show abstract

A Framework for Iris Partial Recognition based on Legendre Wavelet Filter

Cited by 5 publications

References 10 publications

Online Multimodal Compression using Pruning and Knowledge Distillation for Iris Recognition

Online Multimodal Compression using Pruning and Knowledge Distillation for Iris Recognition

Effect of Smaller Fingerprint Sensors on the Security of Fingerprint Authentication

An Efficient Iris Recognition Technique using CNN and Vision Transformer

Contact Info

Product

Resources

About