BrainhackCV

Object Detection is trained on a YOLOv8 model, while reID is trained on a Siamese Network with Triplet Loss

The training and inference code are collated onto one single notebook.

Here are the steps to run it on Colab:

1. Open the notebook on Colab
2. Clone this github repository (https://github.com/wrjgold/BrainhackCV.git)
3. Run all the commands in the notebook for training and inference.

Note:

1. The zip files containing the training, validation, test and suspect data need to be stored in a folder called 'Brainhack' inside your Google Drive. E.g. /content/drive/MyDrive/Brainhack/train.zip
2. You can change the directory path otherwise.