Quantized Neural Networks for Space Applications

Deep learning based approaches can be used for a wide range of space applications such as on-board data processing for observation satellite and collision prevention, spacecraft rendezvous, etc. Unfortunately, the deep learning models are very computational intensive and require huge amount of resources and power consumption. In recent years, some techniques like quantization, pruning and Knowledge distillation, were proposed to compress these deep neural networks. 

In this project, the student will use some recent state of the art methods for quantization of these models. The resulting compressed model, will be evaluated on different benchmarks. The final selection and validation of the selected approach should be optimized to run real-time, on-chip on a specific HW architecture.

Prerequisites:

  • Machine Learning, experience with PyTorch
  • Proficiency in Python
  • Basics of Computer Vision

References:

Contact:

Write to [email protected] and [email protected]