Reproduce “Investigating Adversarial Robustness of Heterogeneous Cooperative Perception” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chenyi Wang, Yutong Liu, Qingzhao Zhang, Ming F. Li
The authors test attacks on cooperative vehicle perception, where collaborators share translated feature maps to detect occluded objects. They introduce HetPoison, an offline-trained, scene-conditioned generator that alters a collaborator’s message in one forward pass. In simulated OPV2V, HEAL-trained HetPoison transfers to STAMP, reducing [email protected] to 0.03 versus 0.681 for same-budget random perturbations; physical vehicles were not tested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.