In computer vision, training a model that performs classification effectively is highly dependent on the extracted features, and the number of training instances. Conventionally, feature detection and extraction are performed by a domainexpert who, in many cases, is expensive to employ and hard to find. Therefore, image descriptors have emerged to automate these tasks. However, designing an image descriptor still requires domain-expert intervention. Moreover, the majority of machine learning algorithms require a large number of training examples to perform well. However, labelled data is not always available or easy to acquire, and dealing with a large dataset can dramatically slow down the training process. In this paper, we propose a novel Genetic Programming based method that automatically synthesises a descriptor using only two training instances per class. The proposed method combines arithmetic operators to evolve a model that takes an image and generates a feature vector. The performance of the proposed method is assessed using six datasets for texture classification with different degrees of rotation, and is compared with seven domain-expert designed descriptors. The results show that the proposed method is robust to rotation, and has significantly outperformed, or achieved a comparable performance to, the baseline methods.

No abstract

The goodness of the features extracted from the instances and the number of training instances are two key components in machine learning, and building an effective model is largely affected by these two factors. Acquiring a large number of training instances is very expensive in some situations such as in the medical domain. Designing a good feature set, on the other hand, is very hard and often requires domain expertise. In computer vision, image descriptors have emerged to automate feature detection and extraction; however, domain-expert intervention is typically needed to develop these descriptors. The aim of this paper is to utilise Genetic Programming to automatically construct a rotation-invariant image descriptor by synthesising a set of formulae using simple arithmetic operators and first-order statistics, and determining the length of the feature vector simultaneously using only two instances per class. Using seven texture classification image datasets, the performance of the proposed method is evaluated and compared against eight domain-expert hand-crafted image descriptors. Quantitatively, the proposed method has significantly outperformed, or achieved comparable performance to, the competitor methods. Qualitatively, the analysis shows that the descriptors evolved by the proposed method can be interpreted.

The incidence of skin cancer, particularly, malignant melanoma, continues to increase worldwide. If such a cancer is not treated at an early stage, it can be fatal. A computer system based on image processing and computer vision techniques, having good diagnostic ability, can provide a quantitative evaluation of these skin cancer cites called skin lesions. The size of a medical image is usually large and therefore requires reduction in dimensionality before being processed by a classification algorithm. Feature selection and construction are effective techniques in reducing the dimensionality while improving classification performance. This work develops a novel genetic programming (GP) based two-stage approach to feature selection and feature construction for skin cancer image classification. Local binary pattern is used to extract gray and colour features from the dermoscopy images. The results of our proposed method have shown that the GP selected and constructed features have promising ability to improve the performance of commonly used classification algorithms. In comparison with using the full set of available features, the GP selected and constructed features have shown significantly better or comparable performance in most cases. Furthermore, the analysis of the evolved feature sets demonstrates the insights of skin cancer properties and validates the feature selection ability of GP to distinguish between benign and malignant cancer images.

© 2017 IEEE. Genetic programming (GP) is a well-known evolutionary computation technique, which has been successfully used to solve various problems, such as optimization, image analysis, and classification. Transfer learning is a type of machine learning approach that can be used to solve complex tasks. Transfer learning has been introduced to GP to solve complex Boolean and symbolic regression problems with some promise. However, the use of transfer learning with GP has not been investigated to address complex image classification tasks with noise and rotations, where GP cannot achieve satisfactory performance, but GP with transfer learning may improve the performance. In this paper, we propose a novel approach based on transfer learning and GP to solve complex image classification problems by extracting and reusing blocks of knowledge/information, which are automatically discovered from similar as well as different image classification tasks during the evolutionary process. The proposed approach is evaluated on three texture data sets and three office data sets of image classification benchmarks, and achieves better classification performance than the state-of-the-art image classification algorithm. Further analysis on the evolved solutions/trees shows that the proposed approach with transfer learning can successfully discover and reuse knowledge/information extracted from similar or different problems to improve its performance on complex image classification problems.