Object detection in multi-modal images using genetic programming

In this paper, we learn to discover composite operators and features that are synthesized from combinations of primitive image processing operations for object detection. Our approach is based on genetic programming (GP). The motivation for using GP-based learning is that we hope to automate the design of object detection system by automatically synthesizing object detection procedures from primitive operations and primitive features. There are many basic operations that can operate on images and the ways of combining these primitive operations to perform meaningful processing for object detection are almost infinite. The human expert, limited by experience, knowledge and time, can only try a very small number of conventional combinations. Genetic programming, on the other hand, attempts many unconventional combinations that may never be imagined by human experts. In some cases, these unconventional combinations yield exceptionally good results. To improve the efficiency of GP, we propose soft composite operator size limit to control the code bloat problem while at the same time avoid severe restriction on the GP search. Our experiments, which are performed on selected regions of images to improve training efficiency, show that GP can synthesize effective composite operators consisting of pre-designed primitive operators and primitive features to effectively detect objects in images and the learned composite operators can be applied to the whole training image and other similar testing images.