Intelligent Multimodal Real Estate Assessment

Innovative ways to tap new sources of real estate related information through multimodal analysis, modeling, and machine learning.

The aim of the project is to develop a method based on multimodal and multitasking machine learning combined with image analysis to extract all relevant information for real estate valuation. The challenge in this is to combine and extract very heterogeneous data and to make them available for further analysis. This includes structured and unstructured text data as well as images and covers both proprietary and non-proprietary data sources (e.g., purchase price contracts from the land registry, images, geographic information, etc.).

The successful modeling of complementary information derived from different data sources (modalities) is key to the robust extraction of real estate related information and to achieving high data quality. To this end, the participating research institutions (University of Applied Sciences Kufstein Tirol and University of Applied Sciences St. Pölten) have launched the interdisciplinary research initiative ImmoPixel, which focuses on novel and innovative ways to unlock new real estate related information sources through multimodal analysis, modeling, and machine learning. The long experience of the academic project partners in combination with the rich multimodal data provided by the project partner Data Science Service (DSS) offers a unique opportunity to take automated real estate valuation to the next level.

Project Partners



Project Manager

Prof. (FH) Dr. Miroslav Despotovic, MA
Professor (FH) for Information Systems for Energy & Real Estate Management