Methods for accuracy‐preserving acceleration of large‐scale comparisons in CPU‐based iris recognition systems

Rathgeb, Christian; Buchmann, Nicolas; Hofbauer, Heinz; Baier, Harald; Uhl, Andreas; Busch, Christoph

doi:10.1049/iet-bmt.2016.0125

Cited by 5 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Binarisation of feature vectors offers an attractive alternative -such templates can be stored efficiently and be compared quickly in the Hamming domain utilising intrinsic CPU operations (i.e. xor and popcount) [7].…”

Section: Introductionmentioning

confidence: 99%

Benchmarking Binarisation Schemes for Deep Face Templates

Drozdowski

Struck

Rathgeb

et al. 2018

2018 25th IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Feature vectors extracted from biometric characteristics are often represented using floating point values. It is, however, more appealing to store and compare feature vectors in a binary representation, since it generally requires less storage and facilitates efficient comparators which utilise intrinsic bit operations. Furthermore, the binary representations are very often necessary for some specific application scenarios, e.g. template protection and indexing. In recent years, usage of deep neural networks for facial recognition has vastly improved the biometric performance of said systems. In this paper, various binarisation schemes are applied to such feature vectors and benchmarked for biometric performance. It is shown that with only a negligible drop in biometric performance, the storage space and computational requirements can be vastly decreased.

show abstract

Section: Introductionmentioning

confidence: 99%

Benchmarking Binarisation Schemes for Deep Face Templates

Drozdowski

Struck

Rathgeb

et al. 2018

2018 25th IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…Lastly, software acceleration and optimisation are also worth mentioning in this context; although there does not seem to be many scientific publications on the topic. In [161], an extensive analysis of possible speed‐ups in CPU‐based iris code comparisons is presented. The authors consider possible improvements through low‐level implementations, manual loop unrolling, caching and pre‐computing certain parts of data, analysis of memory access bottlenecks, multi‐threading, as well as statistical optimisation of micro‐operations.…”

Section: Computational Workload Reduction Approachesmentioning

confidence: 99%

Computational workload in biometric identification systems: an overview

2019

Self Cite

View full text Add to dashboard Cite

Computational workload is one of the key challenges in biometric identification systems. The naïve retrieval method based on an exhaustive search becomes impractical with the growth of the number of the enrolled data subjects. Consequently, in recent years, many methods with the aim of reducing or optimising the computational workload, and thereby speeding-up the identification transactions, in biometric identification systems have been developed. In this article, a taxonomy for conceptual categorisation of such methods is presented, followed by a comprehensive survey of the relevant academic publications, including computational workload reduction and software/hardware-based acceleration. Lastly, the pertinent technical considerations and trade-offs of the surveyed methods are discussed, along with an industry perspective, and open issues/challenges in the field.

show abstract

“…The annular portion among the snowy sclera also dark pupil is called iris and it contains large amount of texture information which is help for the iris identification scheme [8]. The core goal of this anticipated system is to design a system responsible for classifying different iris images.…”

Section: Fig 1 Example Of Iris Recognition At a Distancementioning

confidence: 99%

“…The active contour technique contains pints of controls to move across the image to gain the equilibrium to identify the edge of the eyes [8].…”

Section: Litrature Serveymentioning

confidence: 99%

Iris Recognition at-a-Distance by Means of Chronological MBO-Based DBN

Shirke¹,

Rajabhushnam²

2019

IJITEE

View full text Add to dashboard Cite

Now a days, Iris recognition is wieldy used for the identification of person. The superior bit of 1 countries exploits biometric system for safety reason with the conclusion goal that in runway boarding, custom freedom, gathering passage, etc. The Iris detection at-a-Distance (IAAD) framework is generally used to identify the person in most of the applications. In this system, different features of iris image are extracted in addition enhances the superiority of iris image. Over the span of the most recent ages there consume raised various structures to design and finish iris affirmation systems which works at longer separation going from one meter to sixty meter. Because of such long scope of iris detection schemes in addition iris attainment scheme provides for the best applications to the client. Therefore, It is necessary to design an effective algorithm for IAAD is necessary. In this article, an actual method for iris recognition is presents. A Chronological Monarch Butterfly Optimization -based Deep Belief Network (Chronological MBO-based DBN) technique is anticipated for iris detection.This technique algorithm is the combination of Chronological theory with the Monarch Butterfly Optimization. It is utilized to mastermind the sequential presumption of an iris picture. Additionally, the Hough Transform calculation is utilized for discovery of iris circle and edge. To enhance the accuracy of anticipated iris recognition system ScatT-Loop descriptor and the Local Gradient Pattern (LGP) are fed to the Chronological MBO-based DBN algorithm and these are castoff to abstract the dissimilar features of an iris picture. The dataset used for these tactices are UBIRIS.v1 For the normalization and segmentation of an iris image is done by by means of Dougman's rubber sheet model. This system is established on MATLAB for executing the Hough transform procedures also for reading the iris images. The simulation results shows that this system successfully recognize the iris at a distance 4 to 8 meter. Different performance parameters like as FAR accuracy, too FRR shows better results in this anticipated work.

show abstract

Methods for accuracy‐preserving acceleration of large‐scale comparisons in CPU‐based iris recognition systems

Cited by 5 publications

References 29 publications

Benchmarking Binarisation Schemes for Deep Face Templates

Benchmarking Binarisation Schemes for Deep Face Templates

Computational workload in biometric identification systems: an overview

Iris Recognition at-a-Distance by Means of Chronological MBO-Based DBN

Contact Info

Product

Resources

About