文と画像情報からの相対概念の獲得

概要

論文の詳細を見る
Recent language acquisition systems which efficiently learn the vocabulary and it's meanings by using linguistic and non-linguistic information has been studied. Formerly, we made a concept acquisition system for the purpose of formalizing the method to acquire the concept with visual scene and sentence. Our system acquires concepts without a priori knowledge. The purpose of this study is to acquire the relative concept from sentence and image information. In this study, we assume that each concept about names, locations, colors and size of objects is already known from the method of previous study. We tried to acquire the relative position or relative size between two objects. The algorithm is to store the information about two learning data and to lead the relative relations. Consequently, some relative concepts are acquired by sequencially learning from features of image and its character strings corresponding to sentence. We have realized the first stage of human's concept acquisition process on a computer system.
愛知教育大学の論文
1995-02-10