ProIn: Learning to Predict Trajectory Based on Progressive Interactions for Autonomous Driving

arXiv:2403.16374v1 Announce Type: new
Abstract: Accurate motion prediction of pedestrians, cyclists, and other surrounding vehicles (all called agents) is very important for autonomous driving. Most existing works capture map information through an one-stage interaction with map by vector-based attention, to provide map constraints for social interaction and multi-modal differentiation. However, these methods have to encode all required map rules into the focal agent's feature, so as to retain all possible intentions' paths while at the meantime to adapt to potential …

abstract agents arxiv attention autonomous autonomous driving constraints cs.cv cs.lg cs.ro driving information interactions map modal multi-modal pedestrians prediction social stage through trajectory type vector vehicles