OEM&Lieferant Ausgabe 2/2021

96 Engineering Partner Artificial intelligence in the automotive industry Black box neural network: Interactive visualization improves understanding of decision-making processes in autonomous vehicles By Annika Mahl, ARRK Engineering GmbH, Munich In self-driving cars, reliable Convolutional Neuronal Networks (CNN) are essential. With their help, artificial intel- ligence is supposed to automatically recognize other traffic participants. However, the more autonomously the car drives, the greater the demands on the safety of the algorithms. In order to protect the human life, a deep understanding of the inner-processes of these neural networks is necessary. Yet, a CNN operates as a black box per se, so the complex decision paths are difficult to understand and are making it hard to as- sess any safety risks. These challenges can be solved by appropriate visualization. For this purpose, ARRK Engineering GmbH has developed a validation tool for analyzing the decision-making processes. Thanks to its interactive visualization, the program allows a deeper insight into each layer of a CNN. All weights of the neurons can be manual- ly adjusted to see their impact on the final object recognition as well as the influences of different confounding factors and certain training methods. Image: © shutterstock.com/sdecoret With the help of Convolutional Neuronal Networks (CNN), artificial intelligences are to recognize pedestrians in autonomous driving, for example. However, the process leading up to a CNN‘s final decision remains a complex black box, making the development of standardized validation difficult. Václav Diviš “In the end, our goal is to minimize or better partition the number of critical neurons so that we can rely on robust image recognition from the AI,” reports Václav Diviš, Senior Engineer ADAS & Autonomous Driving at ARRK Engineering. “We are therefore already looking forward to feedback from the field, as this will allow us to further optimize the tool.” Image: © ARRK Engineering GmbH The current methods for analyzing and val- idating neural networks originate primarily from scientific research. However, these methods rarely take into account indus- try-standard functional safety requirements. “Therefore, in order to better understand the issues in object recognition by CNNs, we have developed software that enables standardized validation,” reports Václav Diviš, Senior Engineer ADAS & Autonomous