To provide a solution to this problem, we propose a novel branched network G-CIL for the navigation policy learning.
We suggest that introducing an attention mechanism to infer the importance of different neighbors is critical for accurate trajectory prediction in scenes with varying crowd size.
Fully investigating the social interactions within the crowd is crucial for accurate pedestrian trajectory prediction.
Forecasting human trajectories is critical for tasks such as robot crowd navigation and autonomous driving.
Previous work has shown the power of deep reinforcement learning frameworks to train efficient policies.
Vision-based autonomous driving through imitation learning mimics the behaviors of human drivers by training on pairs of data of raw driver-view images and actions.
The basic idea behind behavioral cloning is to have the neural network learn from observing a human expert's behavior.
End-to-end visual-based imitation learning has been widely applied in autonomous driving.