Satellite 6 Degree of freedom pose estimation
These are notebook for solving the 6D satellite pose estimation challenge proposed by the European Space Agency in the kelvin project cf. https://kelvins.esa.int/pose-estimation-challenge-post-mortem/home/
- Python modules requirements : torch, torchvision, pytorch-lightning, numpy, matplotlib, pillow
You can download SPEED Data from here : https://kelvins.esa.int/satellite-pose-estimation-challenge/data/
- If you want to train locally :
open ./local_runs/local_run_notebook.pynb
change DATA_PATH to the path of your local SPEED data directory after download, cf. #FIRST_TIME_SETUP comment
- If you want use the azure ml training pipline :
open azure_ml_run_notebook.pynb
creat an azure account
change your DATA_PATH, subscription and tenant ids cf. #FIRST_TIME_SETUP comment
if you want to costumize your working space, compute, dataset, environement, and experiment change the corresponding variables along the notebook
Any advise for common problems or issues.
command to run if program contains helper info
Contributors names and contact info
Salem Ben el cadi My linkedin
- 0.2
- Various bug fixes and optimizations
- See commit change or See release history
- 0.1
- Initial Release
This project is licensed under the Apache License - see the LICENSE.md file for details
Inspiration, code snippets, etc.