AI Based Virtual Interior Design

Abstract

In response to the inefficiencies of traditional interior space design classification, this study proposes a deep learning-based approach using a Convolutional Neural Network (CNN) architecture to automate interior design type recognition. Specifically, the NASNetMobile model is employed to classify interior design images into five major categories: Bathroom, Bedroom, Dining, Kitchen, and Living Room. A web-based application was developed using the Flask framework, allowing users to input an interior type and retrieve relevant images. The model was trained on a curated dataset of interior design images and achieved high classification accuracy. Image preprocessing techniques such as resizing, normalization, and augmentation were applied to improve model generalization. The proposed system enhances the speed and accuracy of identifying interior design types, supporting both designers and end users in retrieving contextually appropriate design inspirations. This work demonstrates the potential of combining deep learning with web applications to build practical tools in the domain of computer vision and interior design automation.