Machine Learning is an essential tool for optimising many components in egeniouss.
For instance, it does not only provide the semantic understanding of images but helps to make image descriptors more salient and supports estimating transformations between images in a sequence.
A variety of neural networks are deployed and continuously optimised to improve egeniouss’ performance over time.
egeniouss utilises highly accurate aerial and terrestrial mobile mapping geo data as reference for visual localisation. This data captures the world from different perspectives and hence enables visual localisation for a wide variety of platforms and operating modes.
egeniouss is designed as a cloud solution to facilitate scalability, reliability and performance as well as the management of a large amount of reference data. It also allows for offline operation through caching if no internet connection is available.
egeniouss is made available through a platform-independent API. This enables its convenient deployment to mobile phones and drones but also other devices and platforms in the future.