rizavelioglu
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -17,6 +17,24 @@ The FashionFail dataset, proposed in the paper ["FashionFail: Addressing Failure
|
|
17 |
comprises 2,495 high-resolution images (2400x2400 pixels) of products found on e-commerce websites. The dataset is divided into training, validation, and test sets, consisting of 1,344, 150, and 1,001 images, respectively.
|
18 |
|
19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
### Annotation format
|
21 |
|
22 |
We follow the annotation format of the [COCO dataset](https://cocodataset.org/#format-data). The annotations are stored in the [JSON format](http://www.json.org/) and are organized as follows:
|
|
|
17 |
comprises 2,495 high-resolution images (2400x2400 pixels) of products found on e-commerce websites. The dataset is divided into training, validation, and test sets, consisting of 1,344, 150, and 1,001 images, respectively.
|
18 |
|
19 |
|
20 |
+
### Download Dataset
|
21 |
+
To address concerns regarding data regulations, we share only the URLs of the images, rather than sharing the image files directly.
|
22 |
+
However, we provide a simple script to facilitate dataset construction.
|
23 |
+
The script initially retrieves annotation files from HuggingFace Datasets, then proceeds to download images using the URLs provided in those annotation files.
|
24 |
+
|
25 |
+
First, install the repository with:
|
26 |
+
```
|
27 |
+
git clone https://github.com/rizavelioglu/fashionfail.git
|
28 |
+
cd fashionfail
|
29 |
+
pip install -e .
|
30 |
+
```
|
31 |
+
|
32 |
+
Then, execute the following script with:
|
33 |
+
```
|
34 |
+
python fashionfail/data/make_dataset.py \
|
35 |
+
--save_dir "dir/to/save" \ # [optional] default: "~/.cache/fashionfail/"
|
36 |
+
```
|
37 |
+
|
38 |
### Annotation format
|
39 |
|
40 |
We follow the annotation format of the [COCO dataset](https://cocodataset.org/#format-data). The annotations are stored in the [JSON format](http://www.json.org/) and are organized as follows:
|