Training data for autonomous driving | 2/7/2019 | Staff
Click For Photo:

Autonomous cars must perceive their environment true to reality. The corresponding algorithms are trained using a large number of image and video recordings. For the algorithm to recognize single image elements, such as a tree, a pedestrian or a road sign, these are labeled. Labeling is improved and accelerated by, a startup established by computer scientist Philip Kessler, who studied at Karlsruhe Institute of Technology (KIT), and his co-founder Marc Mengler.

"An algorithm learns by examples and the more examples exist, the better it learns," Philip Kessler says. For this reason, automotive industry needs a large amount of video and image material in machine learning for autonomous driving. So far, objects on the images have been labeled manually by human staff. "Big companies, such as Tesla, employ thousands of workers in Nigeria or India for this purpose. The process is troublesome and time-consuming," Kessler explains. "We at use artificial intelligence to make labeling up to ten times quicker and more precise," he adds. Although image processing is highly automated in large parts, final quality control is made by humans. Combination of technology and human care is particularly important for safety-critical activities, such as autonomous driving," the founder of says. The labelings, also called annotations, in the image and video files have to agree with the real environment with pixel accuracy. The better the quality of the processed image data, the better is the algorithm that uses these data for training.

Training - Images - Situations - Accidents

"As training images cannot be supplied for all situations, such as accidents, we now...
(Excerpt) Read more at:
Wake Up To Breaking News!
Hell sometimes looks an awful lot like an office cubicle.
Sign In or Register to comment.

Welcome to Long Room!

Where The World Finds Its News!