Preprocess the UA-DETRAC dataset and create a dataset for YOLOv5.
- Download the UA-DETRAC dataset from here and unzip it.
- Use
rename_image.py
to rename the images in the dataset. - Use
xml2txt.py
to convert the xml files in the dataset to txt files. - Use
rename_txt.py
to rename the txt files in the dataset. - (Optional)Use
split_data.py
to reorganize the dataset and divide it into training, validation, and testing sets according to a certain proportion. The method I adopted is to keep the original training set unchanged, and the original test set is divided into a new validation set and a test set in a 1:1 ratio.