<![CDATA[The development of autonomous cars is currently increasing along with the need for safe and comfortable autonomous cars. The development of autonomous cars cannot be separated from the use of deep learning to determine the steering angle of an autonomous car according to the road conditions it faces. In the research, a Vision Transformer (ViT) model is proposed to determine the steering angle based on images taken using a front-facing camera on an autonomous car. The dataset used to train ViT is a public dataset. The dataset is taken from streets around Rancho Palos Verdes and San Pedro, California. The number of images is 45,560, which are labeled with the steering angle value for each image. The proposed model can predict steering angle well. Then, the steering angle prediction results are compared using the same dataset with existing models. The experimental results show that the proposed model has better accuracy regarding the resulting MSE value of 2,991 compared to the CNN-based model of 5,358 and the CNN-LSTM combination model of 4,065. From the results of this experiment, the ViT model can replace the existing model, namely the CNN model and the combination model between CNN and LSTM, in predicting the steering angle of an autonomous car.]]>
Copyrights © 2023
