• Stars
    star
    132
  • Rank 274,205 (Top 6 %)
  • Language
  • Created over 5 years ago
  • Updated over 4 years ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

Fashionpedia Dataset

Fashionpedia is a new dataset which consists of two parts: (1) an ontology built by fashion experts containing 27 main apparel categories, 19 apparel parts, 294 fine-grained attributes and their relationships; (2) a dataset with everyday and celebrity event fashion images annotated with segmentation masks and their associated per-mask fine-grained attributes, built upon the Fashionpedia ontology.

Find out more information about Fashionpedia at below links:

Download

CVDF hosts the images and annotations in the Fashionpedia dataset.

Images

Annotations

Detection: apparel object instance segmentation with localized attributes prediction:

Global attributes prediction:

Annotation format

We follow the annotation format of the COCO dataset with additonal fields, such as attributes. The annotations are stored in the JSON format and are organized as follows:

Detection task (instances_attributes)

{
 "info": info,
 "categories": [category],
 "attributes": [attribute],
 "images": [image],
 "annotations": [annotation],
 "licenses": [license]
}

info{
  "year" : int,
  "version" : str,
  "description" : str,
  "contributor" : str,
  "url" : str,
  "date_created" : datetime,
}

category{
  "id" : int,
  "name" : str,
  "supercategory" : str,  # parent of this label
  "level": int,           # levels in the taxonomy
  "taxonomy_id": string,
}

attribute{
  "id" : int,
  "name" : str,
  "supercategory" : str,  # parent of this label
  "level": int,           # levels in the taxonomy
  "taxonomy_id": string,
}

image{
  "id" : int,
  "width" : int,
  "height" : int,
  "file_name" : str,
  "license" : int,
  "time_captured": string,
  "original_url": string,
  "isstatic": int, 0: the original_url is not a static url,
  "kaggle_id": str,
}

annotation{
  "id" : int,
  "image_id" : int,
  "category_id" : int,
  "attribute_ids": [int],
  "segmentation" : [polygon] or [rle]
  "bbox" : [x,y,width,height], # int
  "area" : int
  "iscrowd": int (1 or 0)
}
polygon: [x1, y1, x2, y2, ...], where x, y are the coordinates of vertices, int
rle: {"size", (height, widht), "counts": str}

license{
  "id" : int,
  "name" : str,
  "url" : str
}

Global attribute prediction task (attributes)

{
 "info": info,
 "attributes": [attribute],
 "images": [image],
 "annotations": [annotation],
 "licenses": [license]
}

annotation{
  "image_id" : int,
  "attribute_ids": [int],
}

# other fields follow the same format as detection task