Comparing the Performance of Different Architectures for Predicting Steering Angles of Cars

Accurately predicting the steering angles of cars is a crucial task in the development of autonomous driving systems. Deep learning models have shown significant improvements in this area, with various neural network architectures being proposed for this task. In this paper, we implemented and compared three different models for predicting steering angles from images in the context of autonomous driving. Mean squared error (MSE) and mean absolute error (MAE) are two metrics that were used to assess the performance of the three models.