DriveSeg Dataset for Dynamic Driving Scene Segmentation

From Communauté de la Fabrique des Mobilités


Dataset for Dynamic Driving Scene Segmentation for autonomous driving car

💼 porté par MIT, Toyota

Driveseg.png


DriveSeg contains more precise, pixel-level representations of many of these same common road objects, but through the lens of a continuous video driving scene. This type of full scene segmentation can be particularly helpful for identifying more amorphous objects – such as road construction and vegetation – that do not always have such defined and uniform shapes. The dataset is comprised of two parts:

DriveSeg (Manual)

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 dataset can be downloaded from the IEEE DataPort or demoed as a video.

Technical Summary:

Video data - 2 minutes 47 seconds (5,000 frame) 1080P (1920x1080) 30 fps

Class definitions (12) - vehicle, pedestrian, road, sidewalk, bicycle, motorcycle, building, terrain (horizontal vegetation), vegetation (vertical vegetation), pole, traffic light, and traffic sign

DriveSeg (Semi-auto)

A set of forward facing frame-by-frame pixel level semantic labeled dataset (coarsely annotated through a novel semiautomatic annotation approach developed by MIT) captured from moving vehicles driving in a range of real world scenarios drawn from MIT Advanced Vehicle Technology (AVT) Consortium data.

The dataset can be downloaded from the IEEE DataPort.

Technical Summary:

Video data - Sixty seven 10 second 720P (1280x720) 30 fps videos (20,100 frames)

Class definitions (12) - vehicle, pedestrian, road, sidewalk, bicycle, motorcycle, building, terrain (horizontal vegetation), vegetation (vertical vegetation), pole, traffic light, and traffic sign

This work was done in collaboration with the MIT and Toyota Collaborative Safety Research Center (CSRC).


Organizations using or interested in using the resource:

Contributor(s):

Tags: University of MIT, Toyota, Autonomous, autonome, identification

Categories: Données, Communauté

Theme: Voiture Connectée, Navettes autonomes, Traces de mobilité et des données associées, Urbanisme et ville

Referent:

Challenge: Abaisser les barrières pour innover sur le véhicule, Augmenter les connaissances partagées en cartographie et usages des véhicules et réseaux de transports

Key people to solicit:

Other related common: Autonomous Visualization System AVS, Dataset for autonomous driving, Nobleo Autonomous systems Repositories, Open Air Interface (véhicule connecté et autonome), Transportation Mode Identification

Wealth sought:

Required skills:

Community of interest: Communauté autour des navettes autonomes, Communauté du Logiciel Libre, Communauté autour des données ouvertes

License: MIT

Terms of Service (TOS):

Level of development:

Link to my actions board:

Link to my cloud, wiki, drive…:

Needs:

Next step:

Documentation of the experimentations:



Other informations

List of the actors using or willing of using this common: aucun pour le moment

List of the workshop reports related to this common: