Application of Meta-learning Framework Based on Multiple-Capsule Intelligent Neural Systems in Image Classification

With the rapid development of Internet information technology, image data is explosively growing. How to quickly and effectively acquire and manage these image information has become a research hotspot in the computer field. It is precisely because images can display information intuitively, and people can get the information they need directly from images. Therefore, people are more accustomed to using images as a more commonly used medium to replace large amounts of text to convey information, which is also the main reason for the rapid growth of image information. In the face of such vast image data, if it can not be used reasonably and fully, it is obviously a great waste of resources. If we want to make full use of these resources, we need to organize and manage them effectively. The research of image classification involves many interdisciplinary disciplines such as mathematics, biology, medicine, pattern recognition, machine learning, artificial intelligence and so on. Its development has gone through many stages. Recently, the hot research on deep learning methods has brought new opportunities to image classification. In this paper, a meta-learning framework based on multi-intelligent nervous system is proposed for image classification and recognition. We integrate the multiple-capsule intelligent neural systems to construct the efficient model. The experimental results show that the proposed algorithm has higher image recognition rate.