Training Faster R-CNN on a custom dataset

In the following code, we will train the Faster R-CNN algorithm to detect the bounding boxes around objects present in images. For this, we will work on the same truck versus bus detection exercise from the previous chapter:

Find the following code in the Training_Faster_RCNN.ipynb file in the Chapter08 folder on GitHub at https://bit.ly/mcvp-2e.

1. Download the dataset:

   !pip install -qU torch_snippets
   import os
   %%writefile kaggle.json
    {"username":"XXX", "key":"XXX"}
    !mkdir -p ~/.kaggle
    !cp kaggle.json ~/.kaggle/
    !chmod 600 /root/.kaggle/kaggle.json
    !kaggle datasets download -d sixhky/open-images-bus-trucks/
    !unzip -qq open-images-bus-trucks.zip
    !rm open-images-bus-trucks.zip
   

2. Read the DataFrame containing metadata of information about images and their bounding box, and classes:

   from torch_snippets import *
   from PIL import Image
   IMAGE_ROOT ...