Introduction
The safety levels of smart vehicles (SV) are high in structured conditions such as motorways but not in less structured environments such as pedestrian junctions and mixed traffic environments. The EU-funded SSVPI project will explore pedestrian intention prediction, which is key for safe SV operation. Specifically, the project will develop multi-source and multi-modal algorithms that can predict the intentions of pedestrians under challenging lighting conditions. The results will inform multiple governmental and commercial autonomous driving initiatives.
Research Contents
- Pedestrian crossing intention prediction
- Pedestrian trajectory prediction
- Pedestrian privacy protection
- RGB-thermal fusion
Publications
- X. Zhang*, P. Angeloudis, Y. Demiris. ST CrossingPose: A Spatial-Temporal Graph Convolutional Network for Skeleton-based Pedestrian Crossing Intention Prediction, IEEE Transactions on Intelligent Transportations, 2022, link.
- X. Zhang*, Y. Demiris. Visible and Infrared Image Fusion using Deep Learning, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 8, pp. 10535-10554, 2023, link.
- X. Zhang*, P. Angeloudis, Y. Demiris. Dual-branch Spatio-Temporal Graph Neural Networks for Pedestrian Trajectory Prediction, Pattern Recognition, vol. 142, 2023, link.
- X. Zhang*, Y. Demiris. Self-Supervised RGB-T Tracking with Cross-Input Consistency. arXiv preprint arXiv:2301.11274 (2023), link.
Ethics
We have obtained ethics approval from Imperial College London to capture datasets.
Funding
This project is funded by the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant.
Privacy Notice
Click here to download the Privacy Notice of the SSVPI project.
