I am doing PhD in human iris recognition system using deep learning.
Work completed are.
Human iris recognition using Casia , ubiris, MMU and Polaris dataset.
Local Binary pattern , RLBP, PCA, LTP.
Published paper related to LBP.
Implementation of human iris system using deep learning models such as VGG16/19, inception. Mobilenet, efficient b0,b1,b3 and b021k,b121k,b321k, Nasnet architecture.
VGG16, inception, Nasnet and mobilenet had best performance.
Published paper on performance analysis of all architecture.
Implementation of ensemble model using the best architecture from previous analysis using ML classifiers such as randomforest, logistic regression, decision tree. Stack ensemble, voting, bagging methods used for ensembling.
Work pending is Gamadion pattern on dataset and input the features extracted to my ensemble mode; and check performance.
Publish paper on comparison of performance on all dataset for LBP,RLBP,LTP , PCA and Gammadion.
Publish paper on ensemble modeling .
Thesis writing (Introduction done).
Literature and methodology in process.
April 20 is my deadline.
This is my Structure of Complete Thesis:
Chapter -1: Introduction
Chapter-2: Literature Survey
Chapter-3:
Feature Engineering based approaches for Iris recognition
Chapter-4:
Deep Learning based approaches for Iris recognition
Chapter-5:
Iris recognition in uncontrolled environment: a Deep learning based approaches
Chapter-6:
Performance analysis of Deep Learning approaches for Iris recognition
Chapter-7: Comparative analysis and conclusion.
Future scope