Abstract
This paper shows an analysis and model about a mechanism that makes different media information harmonized. Our purpose is to model subjective interpretation of internal space based on 2d and 3d descriptions. Once a model is established, subjective interpretation of unknown information such as 2d image and 3d object data can be predicted. Statistical analysis is applied to obtain the relationship between subjective responses and 2d living room images and 3d chair models. The system consists of two sub-systems to predict interpreted adjective words of living rooms and chairs, and to retrieve 3d chair models harmonized to a living room from database based on user's interpretation. We establish and evaluate the retrieval system. The accuracy of the retrieval result is 72% in terms of the first candidate of 3d chairs. We found that not only adjective relations but also physical descriptive relations between 2d and 3d data are necessary for harmony between living rooms and chairs.
