In this paper, a new approach for obstacle vehicle path prediction is proposed based on a deep neural network, which is important for Advanced Driver Assistance Systems (ADAS) and autonomous vehicles. In order to analyze sequential sensor data, a Recurrent Neural Network (RNN) is used and the input data for RNN is drawn from three sensors: LIDAR, camera and GPS. These sensor data are obtained experimentally with real vehicles. In addition, deep ensemble is used for robustness of the estimation and acquisition of the uncertainty. The predicted path of the proposed method is continuous and it predicts both short-term and long-term path with a single algorithm. The size of the network model is small, but it shows good performance in predicting future trajectory of obstacle vehicles.

Even though some of the driver assistant systems have been commercialized to provide safety and convenience to the driver, they can be applied for autonomous driving in limited situations such as highways. In this paper, we propose a supervisor agent, which can enhance the driver assistant systems by using deep distributional reinforcement learning. The supervisor agent is trained using end-to-end approach which directly maps both a camera image and LIDAR data into action plan. Because the well-trained network of deep reinforcement learning can lead to unexpected actions, collision avoidance function is added to prevent dangerous situations. In addition, the highway driving case is a stochastic environment with inherent randomness and, thus, its training is performed through the distributional reinforcement learning algorithm, which is specialized for stochastic environment. The optimal action for autonomous driving is selected through the return value distribution. Finally, the proposed algorithm is verified through a highway driving simulator which is implemented by the Unity ML-agents.

Deep learning is a fast-growing field of research, in particular, for autonomous application. In this study, a deep learning network based on various sensor data is proposed for identifying the roads where the vehicle is driving. Long-Short Term Memory (LSTM) unit and ensemble learning are utilized for network design and a feature selection technique is applied such that unnecessary sensor data could be excluded without a loss of performance. Real vehicle experiments were carried out for the learning and verification of the proposed deep learning structure. The classification performance was verified through four different test roads. The proposed network shows the classification accuracy of 94.6% in the test data.

The safety of an automated vehicle requires accurate information of surrounding conditions, because a false sensor output can lead to a fatal accident during driving. Thus, monitoring of abnormalities in every sensor is important for robust perception of the environment. Since it is difficult to obtain anomalous data, it is hard to develop a robust detection algorithm using only a relatively small number of anomalies. In this paper, we propose a data augmentation method for oversampling minority anomalies in lane detection. Using a generative adversarial network that makes the generator learn to estimate the distribution of anomalous data, it generates synthesized minority anomalies. The generated anomalies are used to train an anomaly detection network while minimizing latency for use in real situations. During training, the generated anomalies, with various mixed quality, are sampled differently according to their quality. This helps the detection network to be optimized with better quality data. Experimental result shows that when using the proposed anomaly detection framework for monitoring lane abnormality, it improves the performance by 12% when compared to the vanilla recurrent neural network.

Interaction-aware prediction is a critical component for realistic path planning that prevents automated vehicles from overly cautious driving. It requires to consider internal states of other driver such as driving style and intention, which the automated vehicle cannot directly measure. This paper proposes a probabilistic driver model for response prediction given the planned future actions of automated vehicle. The drivers internal states are considered in an unsupervised manner. The prediction model utilizes mixture density network to estimate future acceleration and yaw-rate profile of interacting vehicles. The proposed method is evaluated by using real-world trajectory data.