Pedestrian intention multi-task identification and trajectory prediction method under view angle of intelligent automobile

Abstract

The invention discloses a pedestrian intention multi-task identification and trajectory prediction method under the view angle of an intelligent automobile, and the method comprises the steps: carrying out the multi-task identification of the pedestrian intention and trajectory prediction according to different kinds of spatio-temporal context information captured in an environment, including visual feature information and non-visual feature information, through a novel neural network architecture, employing a hybrid method, and carrying out the multi-task identification of the pedestrian intention in the view angle of the intelligent automobile; performing joint visual space and dynamic reasoning on each information source by using a feedforward network and a loop architecture, fusing visual information and non-visual information of m historical time steps at T time, classifying current states or actions of pedestrians at the time T, predicting future crossing intentions, outputting actions and intention probabilities at the time T, and obtaining a crossing intention prediction result; the model also predicts a trajectory from time T to time T + n. The method comprehensively considers global spatio-temporal context information of a traffic environment where pedestrians are located, comprises five kinds of visual and non-visual information sources, improves the accuracy of pedestrian crossing intention prediction, and has the advantages of being small in occupied memory amount, high in reasoning speed, complementary in associated task performance and the like.

Description

technical field [0001] The invention belongs to the technical field of intelligent driving, and in particular relates to a method for multi-task recognition of pedestrian intentions and trajectory prediction from the perspective of the self-vehicle of an intelligent vehicle. Background technique [0002] Behavior recognition and intention prediction of road users has always been a very challenging problem for intelligent driving systems in urban environments. Under the current mixed traffic environment conditions, pedestrians in the road traffic system lack corresponding protection equipment and become a vulnerable group among road traffic participants. Intelligent driving vehicles use autonomous or networked sensing methods to achieve peripheral perception within the range of On the basis of pedestrian target recognition, predict its crossing intention and future trajectory, so as to assist intelligent vehicles in intelligent decision-making and path planning, which is of g...

AI Technical Summary

Problems solved by technology

The intention prediction method based on historical trajectory focuses on considering the dynamic factors of pedestrians. The existing pedestrian trajectory prediction methods, such as dynamic Bayesian network, recursive Bayesian filter and multi-layer perceptron, are mostly combined with dynamic systems. Poor chemical performance

Later, some methods based on deep learning and reinforcement learning were also used for trajectory prediction and intention recognition based on historical trajectories, but these methods were basically based on the top-down perspective, because they had to overcome the changing camera viewpoint, occlusion and For other dynamic scenarios, it is still challenging to predict the future trajectory and crossing intention of pedestrians using historical trajectories from the perspective of the vehicle

Images