This project performs object detection on various items of clothing from 191,000 images. As a short description of this dataset, there are 191K training images with 13 different object classes and anywhere between 1 and 8 bounding boxes across the dataset. This data is too large to fit into github memory, so please refer to the link below.
First commit with a new folder structure
https://github.com/switchablenorms/DeepFashion2
All the detailed information about this dataset is at this link.
I am using a Faster RCNN network built in PyTorch and trained on AWS. More details to follow.