Vision SDK

When handling video data, it is key to only transfer, store and post-process data with relevant information. Our system has learned to distinguish traffic behavior from professionals: Drivers. Our Vision SDK sees the full picture – beyond ADAS and autonomous related object classes. An urban traffic scene has a high dimensional state space. To generate meaning, we’re assessing causalties of all present inputs. Based on our federated teaching approach, we are able to determine traffic events from a broad range of scenarios. We derive actionable information and gain superior data based insights.