Is There Still Confidence In Hand-Crafted Feature Extraction Techniques To Use Them In Biometric Systems?

Aounallah, Asma; Bradji, Louardi; Bendjenna, Hakim

doi:10.1109/icrami52622.2021.9585981

Cited by 1 publication

(1 citation statement)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These traits feature low risk of falsification, difficult duplication, and stability, as they are under the skin. Moreover, the availability of a device that is able to acquire two types of modalities at the same time has pushed researchers to fuse these two modalities in an attempt to improve unimodal biometric systems by replacing them with multimodal biometric systems [5]. For this, we use the multispectral palm-print images in this work (with several spectral bands) to design a unimodal/multimodal biometric system.…”

Section: Introductionmentioning

confidence: 99%

Learning-Free Deep Features for Multispectral Palm-Print Classification

Bendjenna

Allah²,

Meraoumia³

2023

csci

View full text Add to dashboard Cite

The feature extraction step is a major and crucial step in analyzing and understanding raw data as it has a considerable impact on the system accuracy. Unfortunately, despite the very acceptable results obtained by many handcrafted methods, they can have difficulty representing the features in the case of large databases or with strongly correlated samples. In this context, we proposed a new, simple and lightweight method for deep feature extraction. Our method can be configured to produce four different deep features, each controlled to tune the system accuracy. We have evaluated the performance of our method using a multispectral palmprint based biometric system and the experimental results, using the CASIA database, have shown that our method has high accuracy compared to many current handcrafted feature extraction methods and many well known deep learning based methods.

show abstract

Section: Introductionmentioning

confidence: 99%