Finetuning PyTorch object detection models for the task of crater detection.

• The notebooks under the training/ directory can either be uploaded to Google Colab, or run locally, in which case, ignore the Colab specific commands, and ensure the required packages are installed.
• The predictor notebook requires makes use of pillow, opencv-python, matplotlib, numpy, torchvision, pytorch and ultralytics.
• Provide the paths to the models and images on which you want to predict in predictor.ipynb. The models can either be trained using the training/ notebooks, or can be downloaded from Releases.