We have completed numerous AI projects in the last years, often in the fields of natural language processing or computer vision. Here we present a selection of those.
The HEGEMON project establishes a sovereign benchmarking framework for AI foundation models and AI applications related to the automated analysis of geoinformation critical to national security.
Learn how dida used AI in combination with depth cameras to measure cows' respiratory rates, to perform health monitoring and detect signs of advancing illness, improving health monitoring and animal welfare in dairy farming.
Our AmoKI project uses machine learning and geospatial data to monitor Germany’s open-pit mines, estimating depth, tracking volume changes, and detecting environmental impacts.
Using various satellite images, a suite of computer vision models and mathematical modeling is applied to detect, segment and analyze tailings with respect to their volume and mineralogical content.
Detect anomalous objects from videos taken by cameras on a train.
AI-based detection of nozzle contamination in industrial 3D printing
We teamed up with Cylad Consulting to analyze time data series collected during a base metal purification process and develop a data-driven solution to optimize process parameters.
We developed a multi-level security system with facial recognition for automatic access control.
We simulate internet traffic and bidding scenarios to predict the reach of advertising campaigns.
AI-supported optical defect detection for semiconductor laser production.
We predict crop types from satellite data to support modern agriculture.
An image segmentation algorithm that supports sustainable city planning.
See how we developed a machine learning software to stop environmental destruction through illegal mines, based on satellite data and object detection.
We automated the detection of certain cloud structures for Deutscher Wetterdienst (DWD).
Creative solutions enabled us to automate the process of planning solar systems.
FewTuRe is a project that aims to employ few-shot learning to automate document processing for SMEs, extracting data from documents with minimal training. In collaboration with the Machine Learning Group at HU Berlin and funded by the Federal Ministry of Education and Research of Germany, dida employs transfer learning and generative AI to handle data scarcity and complex formats, ensuring a scalable, transparent solution for digital transformation.
Given a free form vet appointment reason we extract symptoms, diseases and requested services.
Automatically extract numerical attributes from product descriptions in order to enrich the existing database.
Our software protects tenants against excessive service charge bills.
Our algorithm helps citizens through the bureaucracy of registering a business.
Different methods from the field of NLP helped us to create software that spots errors in rental contracts.
