Computer Vision Use Cases – Development & Operation

24.03. | 10:15 - 10:45 | festival.stage

Early detection of Graffiti on Deutsche Bahn trains is crucial to minimizing their negative effects on the user experience. Delays due to train cleaning compound these operational difficulties and underline the necessity for stable, automatic solutions. Artificial intelligence and machine learning are prime technologies to address these issues, though they are not without their risks and pitfalls. Here we address some of these lessons learned, provide a general introduction to the deep learning models we use and discuss MLOps as an approach to integrate experimental ML development into reliable deployment and operation processes.

Nico Becker (Data Scientist | Alexander Thamm GmbH)
Peco Elenchevski (AI Engineer at | DB Systel)