metadata
dataset_info:
features:
- name: image_id
dtype: int64
- name: image
dtype: image
- name: height
dtype: int64
- name: width
dtype: int64
- name: annotations
list:
- name: id
dtype: int64
- name: image_id
dtype: int64
- name: category_id
dtype: int64
- name: bbox
sequence: int64
- name: segmentation
sequence:
sequence: int64
- name: area
dtype: int64
- name: iscrowd
dtype: int64
splits:
- name: train
num_bytes: 8912943
num_examples: 100
- name: validation
num_bytes: 5140831
num_examples: 30
- name: test
num_bytes: 3428543
num_examples: 30
download_size: 17388632
dataset_size: 17482317
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*