Houses price estimation from images and textual data
- 0 Collaborators
Most existing automatic house price estimation systems rely only on some textual data like its neighborhood area and the number of rooms. The final price is estimated by a human agent who visits the house and assesses it visually. In this project, we automatically estimate the house price by extracting visual features from house photographs and combining them with the house’s textual information. ...learn more
Project status: Published/In Market
Overview / Usage
In this project, we provide a framework for automatic house price estimation based on the visual (images) and textual attributes (metadata) of the house. We also provide the first houses dataset, to our knowledge, that combines both visual and textual attributes to be used for the estimation process.
This work has been published in scitepress library: http://www.scitepress.org/Papers/2016/60407/60407.pdf
Methodology / Approach
We use SIFT feature extractor to extract the features from the houses images and then we combine these features with the textual data to be processed by a multi-layer neural network (NN) to learn the houses prices.
Technologies Used
We developed this software using C++ for the SVM experiments and the Neural Networks toolbox in Matlab.
Repository
https://github.com/emanhamed/Houses-dataset