Autonomous Cars: Vision based Steering Wheel Angle Estimation
This addresses the challenge of developing autonomous car-kits that are not tied to specific vehicle models, potentially reducing data collection burdens and hardware costs.
The paper tackles the problem of model-dependent training data and hardware in autonomous cars by proposing a vision-based system that matches road images with steering wheel images, eliminating the need for steering angle sensors and enabling model-agnostic development.
Machine learning models, which are frequently used in self-driving cars, are trained by matching the captured images of the road and the measured angle of the steering wheel. The angle of the steering wheel is generally fetched from steering angle sensor, which is tightly-coupled to the physical aspects of the vehicle at hand. Therefore, a model-agnostic autonomous car-kit is very difficult to be developed and autonomous vehicles need more training data. The proposed vision based steering angle estimation system argues a new approach which basically matches the images of the road captured by an outdoor camera and the images of the steering wheel from an onboard camera, avoiding the burden of collecting model-dependent training data and the use of any other electromechanical hardware.