Research Associate in Efficient Machine Learning Systems for Earth Observation (Fixed Term)
Department of Computer Science and Technology
Applications are invited for a postdoctoral researcher to work on efficient machine-learning systems for earth observation. The post holder will be part of the Computer Architecture group under the guidance of Prof. Robert Mullins.
The role involves developing new low-power systems and algorithms to significantly enhance the capabilities of onboard AI when used within small satellites. The project is a collaborative effort involving teams at the University of Manchester and University of Southampton and partners at the Alan Turing Institute and HMGCC. The research will seek innovations at both the software and hardware level. The use of a novel memory system and recent advances in low-power machine-learning accelerators will create significant scope to explore new research directions. On-board processing tasks will include object detection, compression, interpretation and various image retrieval tasks. The "Perfect Recollection for Clearer Insight" project is funded by EPSRC.