Published Paper Details:

Accurate cost estimation in interior design is a complex challenge characterized by high-dimensional variables and subjective aesthetic choices. Traditional estimation methods often suffer from significant margins of error due to the non-linear relationship between spatial configurations and material costs. This paper proposes a hybrid computational framework that integrates traditional machine learning (ML) with deep learning for enhanced predictive accuracy. A dual-input pipeline is employed: first, a Convolutional Neural Network (CNN) extracts design features and material styles from 2D architectural blueprints; second, these visual features are fused with structured tabular data--such as square footage, geographical location, and material grades--to feed ensemble learning models, including Random Forest and XGBoost. Experimental results demonstrate that the hybrid framework achieves a Mean Absolute Error of 6.14%, representing a significant improvement over the traditional baseline model. The study concludes that the synergy of computer vision and regression-based ML provides a robust, scalable tool for real-time budgetary forecasting in the architectural and interior design industries.