, , ,

What is autonomous driving, labeling of autonomous driving data

Everyone knows autonomous that self-driving cars are going to run on the road, so it needs a lot of vehicle-road coordination technologies, one of which is very important that it needs a lot of road information markings to guide drivers to drive. This sign generally refers to traffic signs. Some cars can be seen parked on the roadside on the road. At this time, we can use traffic signs to guide drivers to the corresponding location. And the data will also include some dynamic information generated by other cars on the road, such as lane lines, headlights, tires, etc.

What is autonomous driving?

The automatic driving system refers to the fully automated and highly centralized control of the vehicle operation system in which the work performed by the vehicle driver is performed. The automatic driving system has the functions of automatic wake-up, start and sleep, automatic driving, automatic cleaning, automatic parking, automatic door opening and closing, etc., and has various operating modes such as normal operation, degraded operation, and operation interruption. The research and development of self-driving cars is to liberate human hands and bring convenience to daily travel.


Labeling of autonomous driving data:

1. 2D vehicle and pedestrian frame labeling: 2D vehicle and pedestrian frame labeling is widely used in the basic recognition of vehicles and pedestrians, that is, to label vehicles and pedestrians in the picture, and to carry out the test model through the attributes of the frame .

2. Vehicle polygon labeling: Vehicle polygon labeling is used to mark the area and classify the vehicle, which is mainly used to identify the vehicle type.

3. 3D cube labeling: 3D labeling of vehicles in 2D pictures is mainly used to train automatic driving to judge the volume of surrounding cars or overtaking vehicles.

4. 3D radar point cloud annotation: 3D radar point cloud annotation is to mark the position and size of the object through 3D images in the video scene. 3D radar point cloud annotation is mainly used in the construction of autonomous driving virtual reality.

5. Labeling of signs and signal lights: labeling of signs and signal lights is a comprehensive labeling of signs and signal lights hanging on the road. The labeling includes area labeling and semantic labeling, so that automatic driving can drive safely according to traffic rules .

6. Lane line labeling: Lane line labeling is a comprehensive labeling of road ground markings, including classification labeling, area labeling and semantic labeling. It is applied to labeling lane lines in intelligent driving scenarios to enable automatic driving Vehicles can travel according to the rules of the lane lines.

7. Semantic Segmentation: Semantic segmentation is a more extensive type of labeling, which is to segment and label different areas in the picture. These types may be “pedestrians, vehicles, buildings, sky, vegetation, etc. This can be very good The help intelligent driving vehicles identify drivable areas on the road.

8. Video tracking and labeling: Video tracking and labeling refers to tracking and labeling the vehicles driving in the video, and marking the frame according to the frame capture of the picture. The marked photos are then recombined and arranged into video data in order, which can be used to train the automatic driving model .

9. ASR voice transcription: ASR voice transcription is often used in the field of voice assistants in automatic driving systems, which can help drivers better manage and control vehicles.

10. Marking of traveling direction: This is a predictive marking of the forward direction of the marked object. It needs to be marked with frame and direction. It is used to train automatic driving to judge the direction of pedestrians or vehicles, and to avoid pedestrians and vehicles.


Table of Contents