Neural Network Ensembles For Image Identification Using Pareto-optimal Features 2014 IEEE World Congress on Computational Intelligence July 6-11 2014 Beijing, China IEEE Congress on Evolutionary Computation Wissam A. Albukhanajer, Yaochu Jin and Johan A. Briffa Department of Computing Faculty of Engineering and Physical Sciences University of Surrey United Kingdom w.albukhanajer@surrey.ac.uk GU2 7XH www.surrey.ac.uk/computing Monday, 7 th July 2014
In this talk: Introduction : • What is RST invariant? • Trace Transform - Invariant Feature Extraction Evolutionary Trace Transform with Noise • Multi objective optimisation • Pareto front An Ensemble Classifier • An Ensemble with different features • Majority voting Experimental Study • Robustness to Scale • Robustness to RST and Gaussian noise • Robustness to RST and Salt & Pepper noise Summary and Conclusions
Introduction What is RST Invariant Feature? The Rotation, Scale and Translation (RST) Invariants 180 170 160 150 A 140 130 A B 120 0 250 500 750 1000 1250 1500 1750 B
Introduction The Trace Transform (TT) Inspired by Radon transform, proposed by Kadyrov and Petrou [1] Trace Transform is a generalisation of Radon Transform. The functional calculated on the image pixels is not necessary the integral. [1] A. Kadyrov and M. Petrou, "The Trace transform and its applications," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, pp. 811-828, 2001.
The Trace Transform (TT) ( Cont. ) Construction of a Triple feature: Figure: Trace Algorithm. A Triple feature from an Image.
Introduction The Trace Transform (TT) ( Cont. ) (p) Different Trace Transforms can be produced Using (a) different trace functionals : (b) Angle ( ϴ ) (p) (p) Angle ( ϴ ) (d) Angle ( ϴ ) (c) Figure: A Butterfly image (a) and its Trace transforms using different Trace functionals: (b) Gradient, (c) Integral and (d) Standard deviation.
Evolutionary Trace Transform with Noise (ETTN) Multi-Objective Optimisation (MOO) The Objectives are to minimize the following: 1 C k ∑ C K µ Ξ = Ξ k ∑∑ 2 ( Ξ ) = Ξ − µ S k jk C w jk k = 1 j k = 1 = 1 k j 1 K ∑ K Ξ Ξ µ = µ ∑ ( Ξ Ξ ) 2 = µ − µ S k K b k = 1 k = 1 k where: ε : is a small quantity to avoid division by zero. � � : Number of samples in class k Ξ �� :The � �� Triple feature of class � � : Number of classes � � : Mean of all classes of Ξ Triple feature µ Ξ : Mean of class � with Triple feature Ξ k
NSGA-II and Pareto Front Table I NSGA-II Parameter Set-up Figure: Pareto front. Three non-dominated solutions are chosen from the optimal front
Funtctionals Description TABLE II: Triple Features Combinations From TABLE III: Functionals Description Evolutionary Trace Transform Algorithm (ETTN) in Fig. 2 TABLE IV: Pairs of Triple Features Combinations From ETTN Algorithm
Ensemble of classifiers Similar MLP base classifiers with Majority voting: different features as inputs Figure: Structure of the ensemble classifier using Pareto optimal feature set.
Experimental Results Each image subjected to random RST deformation + Noise Figure: Fish database [1]. [1] A. Kadyrov and M. Petrou, "The Trace transform and its applications," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, pp. 811-828, 2001.
Robustness to Scale Only Figure: Robustness to scale only (one test sample per class for each scale factor).
Robustness to Gaussian Noise Figure: Robustness to Gaussian noise of zero mean and standard deviation 2 = 0, 2, 4, 6, 8 and 10, of each approach, (Fish-94 database). Performance are shown when the object is scaled from 1 to 0.3, rotated and translated in a random way, and Gaussian noise has been added to the whole image with standard deviation values corresponds to each figure (one test sample per class for each scale factor). [1] A. Kadyrov and M. Petrou, "The Trace transform and its applications," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23, pp. 811-828, 2001.
Robustness to Gaussian Noise
Robustness to Gaussian Noise
Robustness to Salt and Pepper Noise Figure: Robustness to additive salt & pepper noise with percentage of altered pixels % = 0, 1, 2,...,6 of each approach, (Fish-94 database). Performance are shown when the object is scaled from 1 to 0.3, rotated and translated in a random way, and noise has been added to the whole image with noise levels corresponds to each figure (one test sample per class for each scale factor).
Robustness to Salt and Pepper Noise
Robustness to Salt and Pepper Noise
Summary and Conclusions • We presented an ensemble classifier using a set of Pareto-optimal Trace transform features. • While the traditional Trace transform that uses thousands of features, the single classifiers or ensemble classifiers using the features extracted by the evolutionary multi-objective Trace transform (ETTN) are able to accurately classify noisy RST deformed images with a much lower computational cost. • Our results indicate that different Pareto-optimal features can introduce diversity in the ensemble classifier. As a result, no particular effort is needed to generate diverse base classifiers.
Publications • W. A. Albukhanajer, Y. Jin, J. Briffa and G. Williams “Evolutionary Multi-Objective Optimization of Trace Transform for Invariant Feature Extraction” IEEE Congress on Evolutionary Computation CEC, Brisbane, Australia, June 10-15, 2012. • W. A. Albukhanajer, Y. Jin, J. Briffa and G. Williams “A comparative study of multi-objective evolutionary trace transform methods for robust feature extraction,” in Evolutionary Multi- Criterion Optimization, ser. Lecture Notes in Computer Science, R. Purshouse, P. Fleming, C. Fonseca, S. Greco, and J. Shaw, Eds. Springer Berlin Heidelberg, 2013, vol. 7811, pp. 573–586. [Online]. Available: http://dx.doi.org/10.1007/978-3-642-37140-0 43 • W. A. Albukhanajer, J. A. Briffa, and Y. Jin, “Evolutionary Multiobjective Feature Extraction in the Presence of Noise,” Submitted for publication, IEEE Trans. SMC Part B. January 2014. • W. A. Albukhanajer, Y. Jin, and J. A. Briffa, “Neural Network Ensembles for Image Identification Using Pareto-optimal Features,” in 2014 IEEE Congress on Evolutionary Computation (CEC), Beijing, China, 06-11, July 2014. • W. A. Albukhanajer, J. A. Briffa, and Y. Jin, “Classifier Ensembles Using Pareto Optimal Image Features ” To be submitted, IEEE Trans. Image Proc. June 2014.
Q&A Wissam A. Albukhanajer Thank you w.albukhanajer@surrey.ac.uk Nature Inspired Computing and Engineering T: +44(0)1483 68 6059 F: +44(0)1483 68 6051 Department of Computing Faculty of Engineering & Physical Sciences University of Surrey Guildford, UK. GU2 7XH http://www.surrey.ac.uk/computing/ 7 th July 2014
Recommend
More recommend
