This project aims to…


  • Transfer Deep Learning and Cloud Computing solutions to Earth Observation systems considering operational constraints
  • Provide massive image analysis and methods for complex mission planning
  • Foster emergence of novel and efficient applications, improving ground segment and on-board autonomy

Key Results

Reinforcement Learning approach for mission scheduling optimization considering weather forecast uncertainty

Reinforcement Learning : Learning the Best Next Action for a State Machine

Formal validation of satellite commands for onboard autonomy

Unsupervised Learning methods for image analysis :

Boat detection
Change detection
Image retrieval 
Land cover

Members & Partners