Automatically Evolving Rotation-invariant Texture Image Descriptors by Genetic Programming

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 domain-expert 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.

A Reflective Encounter With the Fine Sand Area in a Nursery School Setting

This article draws on a model of reflection that involves creating meanings through repeated encounters with evocative objects. Responses to one such evocative object, a 20-second video clip of children playing in the fine sand area, illustrates the “turning toward” and then “turning away” from the object to engage with broader themes. Parten’s play types are used when analyzing children’s play in the fine sand area (the evocative object). The focus then turns away to themes of English as a second language, messy play, energies of childhood, and Foucault’s docile bodies. The intention was to integrate loosely formulated research aims relating to quality of the physical environment within the evolving life of the nursery setting to encourage a developing research orientation and reflective dis-position. This way of approaching practitioner research is well adapted to a longer-term engagement with enduring areas of interest, such as developing the potential of the physical environment.