Earth observation

Monitoring and prediction of objects, hazard risks and streamlining of aerial surveys


Optical images from drones or satellites and data captured by radar sensors from above contain enormous amounts of complex data. They have the potential to reveal valuable information about our planet and its surface that could be used automate terrain mapping or to predict objects and hazard risks such as vessels and potential oil spills at sea.

Main objective

For earth observation the planned innovations aim for improved methods for monitoring and prediction of hazard risks, object detection, and for surveying and mapping ground and sea from air through exploitation of remote sensing images from satellites, aircrafts and drones.


Limited and inadequate training data is a general problem in remote sensing. Combination of multi-sensor data (e.g. from optical and radar sensors) and time dependencies is another key challenge. Modelling of contextual information may also enhance the performance, but important contextual issues like integration of physical properties have not yet been addressed.

These are some of the research challenges Visual Intelligence are addressing.

Highlighted publications

Addressing Distribution Shifts in Federated Learning for Enhanced Generalization Performance
June 4, 2023
Training and test data from different clients pose a challenge.
Hubs and Hyperspheres: Reducing Hubness and Improving Transductive Few-shot Learning with Hyperspherical Embeddings
March 6, 2023
We approach the representation learning task by tackling the hubness problem.
On the Effects of Self-supervision and Contrastive Alignment in Deep Multi-view Clustering
March 6, 2023
We propose DeepMVC – a unified framework which includes many recent methods as instances.