Pedestrian Trajectory Prediction Method Based on Spatiotemporal Attention Mechanism

Abstract

The pedestrian trajectory prediction method based on the spatiotemporal attention mechanism includes: collecting image information in the scene and extracting the position information of pedestrians in the image coordinate system; performing data preprocessing to obtain the historical trajectory coordinates of each pedestrian in the scene; using an encoder Encoder encodes the pedestrian's historical trajectory and outputs a feature tensor; and uses the decoder Decoder to iteratively predict the pedestrian's future trajectory coordinates; among them, the encoder Encoder uses the attention mechanism to fuse each pedestrian's own historical trajectory information and the same scene. Interactive information between different pedestrians; realizes real-time and effective prediction of pedestrian trajectories in actual autonomous driving application scenarios, which not only adapts to the processing capabilities of the vehicle-mounted low-power processor, but also improves the accuracy of pedestrian trajectory prediction, providing a basis for actual autonomous driving. It provides a reliable basis for driving decisions and greatly improves the safety of autonomous driving.

Description

technical field [0001] The disclosure relates to the technical fields of automatic driving and computer vision, in particular to a pedestrian trajectory prediction method, device, electronic device and storage medium based on a spatio-temporal attention mechanism, and in particular to a method based on deep learning in complex pedestrian interaction scenarios Pedestrian trajectory prediction technology. Background technique [0002] With the development of computer vision technology, the use of computer vision technology for environmental perception has become an indispensable part of automatic driving systems and other intelligent perception systems. Among them, pedestrian trajectory prediction is of great significance in the fields of automatic driving and video surveillance. In the autonomous driving scenario, predicting the future trajectory of pedestrians can assist autonomous vehicles to make correct decisions, ensure the safety of pedestrians, and improve the safety ...

AI Technical Summary

Problems solved by technology

The mainstream methods generally use recurrent neural networks for time series prediction, including RNN, LSTM, GRU, etc. Typical methods such as Social-LSTM cannot be parallelized, and the calculation efficiency is low. It is difficult to model long-distance relationships, which is easy to cause For the technical problem of performance bottlenecks, this disclosure is based on the attention mechanism, which can effectively capture the key parts of the historical trajectories of pedestrians, and can guarantee performance in a parallelized style and global receptive field

Images