Catalogues List

Name: ApolloScape

Domain: AD

Description: ApolloScape, a part of Apollo, additionally offers training data for semantic segmentation (pixel-level classification of video frames, usually input for training neural networks). As of April 2021, the dataset contained 100k video frames, 80k LiDAR point clouds and trajectories covering 1000 km in urban traffic. ApolloScape also includes a scene parsing dataset covering almost 150k frames with corresponding pixel-level annotations and pose information, depth maps for static background.

When: 2019

Where: China

How:

What:

Data Categories: 3D Lidar point clouds, video, trajectories, GNSS

Data Format: Bin and txt

Lables: Different types of high qualiaty labeles available.

Size:

Hours:

Link: https://apolloscape.auto/tracking.html

License: For academic purposes only. User license at https://apolloscape.auto/license.html

Doi:

Accessed: 2024-01-12

Name: Boreas

Domain: AD

Description: Intended to support odometry, metric localization, and 3D object detection for lidar, radar, and vision over a long period and multiple weather conditions

When: 2021

Where: Toronto, Canada

How: 1x Cam, 1x Lidar, 1xGNSS, 1xIMU, 1xRadar

What:

Data Categories: Lidar, radar, video, GNSS and meta data

Data Format: Point cloud data: binary, camera: png images, pose files: csv, metadata: txt and yaml

Lables: N/A

Size:

Hours:

Link: https://www.boreas.utias.utoronto.ca

License: CC Attribution

Doi:

Accessed: 2024-01-22

Name: CarlaScenes

Domain: AD

Description: Synthetic dataset generated in CARLA focused in odometry related tasks.

When: 2022

Where: CARLA (CAR Learning to Act) is an open simulator for urban driving, developed as an open-source layer over Unreal Engine 4.

How: Camera, Semantic Camera, Depth, Lidar, Semantic Lidar, IMU, GNSS.

What:

Data Categories: Point cloud data, video data, metadata

Data Format:

Lables: Ground truth simulation

Size:

Hours:

Link: https://github.com/CarlaScenes/CarlaSence

License:

Doi:

Accessed: 2024-01-24

Name: MIT DriveSeg (Manual) Dataset

Domain: AD

Description: MIT DriveSeg (Manual) Dataset is a forward facing frame-by-frame pixel level semantic labeled dataset captured from a moving vehicle during continuous daylight driving through a crowded city street. The datset is densely annotated for every pixel and every one of 5,000 video frames. The purpose of this dataset is to allow for exploration of the value of temporal dynamics information for full scene segmentation in dynamic, real-world operating environments. The dataset is produced from the MIT-AVT dataset.

When: 2020

Where: US

How:

What:

Data Categories: Images, labels

Data Format: Images: png, labels: not known

Lables: Vehicle, pedestrian, road, sidewalk, bicycle, motorcycle, building, terrain (horizontal vegetation), vegetation (vertical vegetation), pole, traffic light, and traffic sign

Size: 3 GB

Hours:

Link: https://ieee-dataport.org/open-access/mit-driveseg-manual-dataset

License: CC Attribution

Doi: http://dx.doi.org/10.21227/mmke-dv03

Accessed: 2024-01-24

Name: MIT-AVT Clustered Driving Scene Dataset

Domain: AD

Description: The MIT-AVT Clustered Driving Scene Dataset is extracted has been generated from the MIT-AVS naturalistic driving study (https://ieeexplore.ieee.org/document/8751968). The data has been labelled by both manual and automatic annotations. 1,156,592 10-second clips including 450 clusters of common scenes and 5601 edge cases.

When: 2021

Where: US

How: 23 Tesla Model S and Model X vehicles, 2 Volvo S90 vehicles, 2 Range Rover Evoque, and 2 Cadillac CT6 vehicles for both long-term (over a year per driver) and medium-term (one month per driver) naturalistic driving data collection. The recorded data streams include IMU, GPS, and CAN messages, and high-definition video streams of the driver's face, the driver cabin, the forward roadway, and the instrument cluster (on select vehicles). In 2019, 122 participants, 15610 days of participation, 511638 mi, and 7.1 billion video frames. The collected data cover all year, different weather conditions, and different time of day.

What:

Data Categories: Video, metadata, labels

Data Format:

Lables: Weather, Lanes, Illuminance

Size: 4 TB

Hours: 3212

Link: https://doi.org/10.1109/IV47402.2020.9304677

License:

Doi:

Accessed: 2024-01-24

Name: NuPlan

Domain: AD

Description: Large-scale planning benchmark for autonomous driving. NuScene dataset continuation, with 1500 hours of driving scenes and a toolkit for measuring performance of planning techniques. Full release in Q1 2022+C9, Second sub-release available.

When: 2021

Where: Boston, Las Vegas, Pittsburgh, USA, and Singapore

How: 5x LIDAR, 8x camera, IMU, GPS

What:

Data Categories: Lidar, video, GPS, map data and meta data

Data Format:

Lables: Automatically annotated (ML) bounding boxes - vehicles - cycles - pedestrians - traffic cones - barriers - construction zone sign - generic J3(animals, poles, pushable objects

Size: 3 GB

Hours: 1200

Link: https://www.nuscenes.org/nuplan

License: CC Attribution. Free of charge for non-commercial use. Exceptions for third party data.

Doi:

Accessed: 2024-01-23

Name: V2X-SIM

Domain: AD

Description: Synthetic Dataset generated in CARLA with synchronized data of multiple vehicles and RSUs.

When: 2022

Where: CARLA (CAR Learning to Act) is an open simulator for urban driving, developed as an open-source layer over Unreal Engine 4.

How: V2XSet is collected using OpenCDA, which is the first open co-simulation-based research/engineering framework integrated with prototype cooperative driving automation pipelines as well as regular automated driving components (e.g., perception, localization, planning, control). V2X-ViT is build upon OpenCOOD, which is the first Open Cooperative Detection framework for autonomous driving.

What:

Data Categories: Point cloud data, video data, metadata

Data Format: Point cloud data: pcd, video data: png, metadata: yaml, lables: yaml

Lables: Ground truth labels

Size: 100 GB

Hours: 52 sequences

Link: https://github.com/DerrickXuNu/v2x-vit

License: CC Attribution

Doi:

Accessed: