SuperPixel based mid-level image description for image recognition

This study proposes a mid-level feature descriptor and aims to validate improvement on image classification and retrieval tasks. In this paper, we propose a method to explore the conventional feature extraction techniques in the image classification pipeline from a different perspective where mid-level information is also incorporated in order to obtain a superior scene description. We hypothesize that the commonly used pixel based low-level descriptions are useful but can be improved with the introduction of mid-level region information. Hence, we investigate superpixel based image representation to acquire such mid-level information in order to improve the accuracy. Experimental evaluations on image classification and retrieval tasks are performed in order to validate the proposed hypothesis. We have observed a consistent performance increase in terms of Mean Average Precision (MAP) score for different experimental scenarios and image categories.