Cyber2A Workshop: Toy Model for RTS Training and Inference

This repository demonstrates a simplified example of training and running inference on a toy model using the Retrogressive Thaw Slumps (RTS) dataset. It is adapted from the official PyTorch Vision Tutorial.

The copyright for the tutorial content belongs to PyTorch. © Copyright 2024, PyTorch.

Requirements

To set up the environment, I recommend using Conda. You can create the environment using the provided requirements_conda.txt file:

conda create --name <env_name> --file requirements_conda.txt

A requirements.txt file is also provided for pip installation, though it has not been fully tested:

Usage

1. Training the Model

To train the model, run the following command in your terminal:

python train.py

This command trains a Mask R-CNN model with a ResNet-50 backbone using the RTS dataset.

2. Running Inference

To perform inference using the trained model, run:

python inference.py --image-path <path_to_image>

Replace <path_to_image> with the path to your image file.

Pre-trained Model

If training cannot be completed due to time constraints, a pre-trained model is available. You can download the model weights from this link and place the rts_model.pth file in the root directory of the repository.

Notes

• This model is for demonstration purposes and has not been optimized for performance.