ImmoEye is a deep learning system that can automatically extract and harness information from images. Our application automatically sorts images of real estate according to their content.
ImmoEye searches images of real estate for their visible content. The goal is to recognize and classify the image content. In kitchens, the software should additionally determine the year of construction of the installations.
Of particular interest was the question of how efficiently and precisely ImmoEye classifies the components after training.
- Fast: ImmoEye needs less than 2 minutes to automatically classify 1000 images.
- Accurate: ImmoEye correctly classifies more than 85% of the images.
- Adaptable: ImmoEye is expandable and can be used for other image processing tasks with relatively little effort, for example, to assess the condition or standard of a building or for tasks outside the real estate industry.