scene_parse150

Scene parsing is to segment and parse an image into different image regions associated with semantic categories, such as sky, road, person, and bed. MIT Scene Parsing Benchmark (SceneParse150) provides a standard training and evaluation platform for the algorithms of scene parsing.

URL: http://sceneparsing.csail.mit.edu/
DatasetBuilder: tfds.image.scene_parse_150.SceneParse150
Version: v1.0.0
Versions:
- 1.0.0 (default):
Size: 936.97 MiB

Features

FeaturesDict({
    'annotation': Image(shape=(None, None, 3), dtype=tf.uint8),
    'image': Image(shape=(None, None, 3), dtype=tf.uint8),
})

Statistics

Split	Examples
ALL	22,210
TRAIN	20,210
TEST	2,000

Homepage

http://sceneparsing.csail.mit.edu/

Supervised keys (for `as_supervised=True`)

(u'image', u'annotation')

Citation

@inproceedings{zhou2017scene,
title={Scene Parsing through ADE20K Dataset},
author={Zhou, Bolei and Zhao, Hang and Puig, Xavier and Fidler, Sanja and Barriuso, Adela and Torralba, Antonio},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
year={2017}
}