Object Detection is the Smart Labeling model which performs Object detection tasks whose type is Binding Box Labeling.

It can be used when the task type is set to the Bounding box in the project's labeling settings.

Vehicle/Pedestrian

1. Definition of Class

• Vehicle
  Standard vehicles are Sedan or SUV type (except vans, trucks, buses, and special vehicles, excluding vehicle accessories)
• Pedestrian
  Child, adult pedestrian

2. Minimum inference size

• Vehicle : 20px(width) * 25px(height)
• Pedestrian : 10px(width) * 15px(height)

Face/License Plate (non-identification)

1. Definition of Class

• Face
  The face of a pedestrian (including the face of a pedestrian whose shape of the face cannot be identified)
• LicensePlate
  Vehicle registration plate (including the license plate of the vehicle whose contents cannot be identified)

2. Minimum inference size

• Face : 10px(width) * 10px(height)
• LicensePlate : 10px(width) * 10px(height)

Wheel

1. Definition of Class

• Wheel: Side of the wheel where the wheel can be seen

2. Minimum inference size

• Wheel : 15px(width) * 15px(height)

Traffic Sign/Traffic Light

1. Definition of Class

• Traffic Sign: Standard vehicles of Sedan or SUV type (except vans, trucks, buses, and special vehicles)
• Traffic Light: Child, adult pedestrian

2. Minimum inference size

• TrafficSign : 10px(width) * 10px(height)
• TrafficLight : 15px(width)

The extent of the performance guarantee of Smart Labeling

1. Information about images

• Resolution: HD or higher
• Camera lens view angle: 120° or lessHowever, inference performance will be degraded for images that have not been corrected for distortion. e.g., fisheye lens

2. Threshold

• The performance will be degraded for inference instances that are less than threshold 0.8.

3. Exceptions

• Inference performance will be degraded for images that blur occurs in images because of rain, snow, foreign objects, etc.
• Inference performance will be degraded for images that have severe backlighting.

If you have any other inquiries, please get in touch with us at [email protected]