Traffic Signs YOLO3 pytorch

YOLO3 to detect the traffic sign in pytoch. I used the Traffic Signs dataset and convert the given annoations into YOLO fromat. Traffic Signs is the dataset of German Traffic Sign Detection Benchmark datset http://benchmark.ini.rub.de It contains 900 image files with natural pictures 00000.ppm - 00899.ppm. Dataset contains the images files and gt.txt to define define the ImageID, XMin, YMin, XMax, YMax with calls numbers as annotation file.

Default Label

Dataset has annotated with following labels

• 0 = speed limit 20 (prohibitory)
• 1 = speed limit 30 (prohibitory)
• 2 = speed limit 50 (prohibitory)
• 3 = speed limit 60 (prohibitory)
• 4 = speed limit 70 (prohibitory)
• 5 = speed limit 80 (prohibitory)
• 6 = restriction ends 80 (other)
• 7 = speed limit 100 (prohibitory)
• 8 = speed limit 120 (prohibitory)
• 9 = no overtaking (prohibitory)
• 10 = no overtaking (trucks) (prohibitory)
• 11 = priority at next intersection (danger)
• 12 = priority road (other)
• 13 = give way (other)
• 14 = stop (other)
• 15 = no traffic both ways (prohibitory)
• 16 = no trucks (prohibitory)
• 17 = no entry (other)
• 18 = danger (danger)
• 19 = bend left (danger)
• 20 = bend right (danger)
• 21 = bend (danger)
• 22 = uneven road (danger)
• 23 = slippery road (danger)
• 24 = road narrows (danger)
• 25 = construction (danger)
• 26 = traffic signal (danger)
• 27 = pedestrian crossing (danger)
• 28 = school crossing (danger)
• 29 = cycles crossing (danger)
• 30 = snow (danger)
• 31 = animals (danger)
• 32 = restriction ends (other)
• 33 = go right (mandatory)
• 34 = go left (mandatory)
• 35 = go straight (mandatory)
• 36 = go right or straight (mandatory)
• 37 = go left or straight (mandatory)
• 38 = keep right (mandatory)
• 39 = keep left (mandatory)
• 40 = roundabout (mandatory)
• 41 = restriction ends (overtaking) (other)
• 42 = restriction ends (overtaking (trucks)) (other)

Modified Label

Among of 42 labels of annotation i have make it folling lables

• 1 = prohibitory
• 2 = danger
• 3 = mandatory
• 4 = other

Reference

Udemy.com Instructor : Valentyn Sichkar Computer Vision, Machine Learning, Image Processing