jhgan00
commented
1 year ago You can just download the datasets (CUB, In-Shop, and SOP) from the following links and extract them to any path you want:
I think there is no further preprocessing steps other than downloading and extracting the data.
For example, my /data directory looks like this:
/data/
├── CUB_200_2011
│ ├── CUB200(2011)
│ ├── README
│ ├── archive.zip
│ ├── attributes
│ ├── bounding_boxes.txt
│ ├── classes.txt
│ ├── image_class_labels.txt
│ ├── images
│ ├── images.txt
│ ├── parts
│ └── train_test_split.txt
├── In-shop
│ ├── img
│ ├── img.zip
│ └── list_eval_partition.txt
├── Stanford_Online_Products
│ ├── Ebay_final.txt
│ ├── Ebay_info.txt
│ ├── Ebay_test.txt
│ ├── Ebay_train.txt
│ ├── LICENSE
│ ├── README
│ ├── bicycle_final
│ ├── bicycle_final.txt
│ ├── cabinet_final
│ ├── cabinet_final.txt
│ ├── chair_final
│ ├── chair_final.txt
│ ├── coffee_maker_final
│ ├── coffee_maker_final.txt
│ ├── fan_final
│ ├── fan_final.txt
│ ├── kettle_final
│ ├── kettle_final.txt
│ ├── lamp_final
│ ├── lamp_final.txt
│ ├── mug_final
│ ├── mug_final.txt
│ ├── sofa_final
│ ├── sofa_final.txt
│ ├── stapler_final
│ ├── stapler_final.txt
│ ├── table_final
│ ├── table_final.txt
│ ├── toaster_final
│ └── toaster_final.txt
...
TeresaTsang
mennanyang
As there is no "/data" directory, would you mind sharing this part or the data preprocessing process? Thanks!