waymo_open_dataset

The Waymo Open Dataset is comprised of high resolution sensor data collected by Waymo self-driving cars in a wide variety of conditions. This data is licensed for non-commercial use.

FeaturesDict({
    'camera_FRONT': FeaturesDict({
        'image': Image(shape=(1280, 1920, 3), dtype=uint8),
        'labels': Sequence({
            'bbox': BBoxFeature(shape=(4,), dtype=float32),
            'type': ClassLabel(shape=(), dtype=int64, num_classes=5),
        }),
    }),
    'camera_FRONT_LEFT': FeaturesDict({
        'image': Image(shape=(1280, 1920, 3), dtype=uint8),
        'labels': Sequence({
            'bbox': BBoxFeature(shape=(4,), dtype=float32),
            'type': ClassLabel(shape=(), dtype=int64, num_classes=5),
        }),
    }),
    'camera_FRONT_RIGHT': FeaturesDict({
        'image': Image(shape=(1280, 1920, 3), dtype=uint8),
        'labels': Sequence({
            'bbox': BBoxFeature(shape=(4,), dtype=float32),
            'type': ClassLabel(shape=(), dtype=int64, num_classes=5),
        }),
    }),
    'camera_SIDE_LEFT': FeaturesDict({
        'image': Image(shape=(886, 1920, 3), dtype=uint8),
        'labels': Sequence({
            'bbox': BBoxFeature(shape=(4,), dtype=float32),
            'type': ClassLabel(shape=(), dtype=int64, num_classes=5),
        }),
    }),
    'camera_SIDE_RIGHT': FeaturesDict({
        'image': Image(shape=(886, 1920, 3), dtype=uint8),
        'labels': Sequence({
            'bbox': BBoxFeature(shape=(4,), dtype=float32),
            'type': ClassLabel(shape=(), dtype=int64, num_classes=5),
        }),
    }),
    'context': FeaturesDict({
        'name': Text(shape=(), dtype=string),
    }),
    'timestamp_micros': int64,
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
camera_FRONT FeaturesDict
camera_FRONT/image Image (1280, 1920, 3) uint8
camera_FRONT/labels Sequence
camera_FRONT/labels/bbox BBoxFeature (4,) float32
camera_FRONT/labels/type ClassLabel int64
camera_FRONT_LEFT FeaturesDict
camera_FRONT_LEFT/image Image (1280, 1920, 3) uint8
camera_FRONT_LEFT/labels Sequence
camera_FRONT_LEFT/labels/bbox BBoxFeature (4,) float32
camera_FRONT_LEFT/labels/type ClassLabel int64
camera_FRONT_RIGHT FeaturesDict
camera_FRONT_RIGHT/image Image (1280, 1920, 3) uint8
camera_FRONT_RIGHT/labels Sequence
camera_FRONT_RIGHT/labels/bbox BBoxFeature (4,) float32
camera_FRONT_RIGHT/labels/type ClassLabel int64
camera_SIDE_LEFT FeaturesDict
camera_SIDE_LEFT/image Image (886, 1920, 3) uint8
camera_SIDE_LEFT/labels Sequence
camera_SIDE_LEFT/labels/bbox BBoxFeature (4,) float32
camera_SIDE_LEFT/labels/type ClassLabel int64
camera_SIDE_RIGHT FeaturesDict
camera_SIDE_RIGHT/image Image (886, 1920, 3) uint8
camera_SIDE_RIGHT/labels Sequence
camera_SIDE_RIGHT/labels/bbox BBoxFeature (4,) float32
camera_SIDE_RIGHT/labels/type ClassLabel int64
context FeaturesDict
context/name Text string
timestamp_micros Tensor int64
@InProceedings{Sun_2020_CVPR,
author = {Sun, Pei and Kretzschmar, Henrik and Dotiwalla, Xerxes and Chouard, Aurelien and Patnaik, Vijaysai and Tsui, Paul and Guo, James and Zhou, Yin and Chai, Yuning and Caine, Benjamin and Vasudevan, Vijay and Han, Wei and Ngiam, Jiquan and Zhao, Hang and Timofeev, Aleksei and Ettinger, Scott and Krivokon, Maxim and Gao, Amy and Joshi, Aditya and Zhang, Yu and Shlens, Jonathon and Chen, Zhifeng and Anguelov, Dragomir},
title = {Scalability in Perception for Autonomous Driving: Waymo Open Dataset},
booktitle = {The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}

waymo_open_dataset/v1.2 (default config)

  • Config description: Waymo Open Dataset v1.2

  • Dataset size: 336.62 GiB

  • Splits:

Split Examples
'train' 158,081
'validation' 39,987

waymo_open_dataset/v1.1

  • Config description: Waymo Open Dataset v1.1

  • Dataset size: 336.62 GiB

  • Splits:

Split Examples
'train' 158,081
'validation' 39,987

waymo_open_dataset/v1.0

  • Config description: Waymo Open Dataset v1.0 This dataset is also available in pre-processed format, making it faster to load, if you select the correct data_dir:
tfds.load('waymo_open_dataset/v1.0', data_dir='gs://waymo_open_dataset_v_1_0_0_individual_files/tensorflow_datasets')
  • Dataset size: 34.73 GiB

  • Splits:

Split Examples
'train' 14,884
'validation' 4,954