Fooling Neural Networks for Motion Forecasting via Adversarial Attacks

arXiv:2403.04954v1 Announce Type: new
Abstract: Human motion prediction is still an open problem, which is extremely important for autonomous driving and safety applications. Although there are great advances in this area, the widely studied topic of adversarial attacks has not been applied to multi-regression models such as GCNs and MLP-based architectures in human motion prediction. This work intends to reduce this gap using extensive quantitative and qualitative experiments in state-of-the-art architectures similar to the initial stages of adversarial attacks in …

abstract advances adversarial adversarial attacks applications architectures arxiv attacks autonomous autonomous driving cs.ai cs.cv driving forecasting human mlp networks neural networks prediction regression safety type via