An image retrieval system to estimate impression words from images using a neural network

摘要

In this paper, we propose a new method to estimate impression words from images using neural network and the image retrieval system. Since the human sensitivity and various images are flexibly connected using a neural network in the proposed system, it can automatically estimate impression words from various kinds of images. As a result, the proposed system can reflect human's kansei (impression and sensitivity) effectively and treat wider variety of images. We show some experimental results. The proposed system can estimate correct impression words and successfully retrieve correct images.