BadFusion: 2D-Oriented Backdoor Attacks against 3D Object Detection

arXiv:2405.03884v1 Announce Type: new
Abstract: 3D object detection plays an important role in autonomous driving; however, its vulnerability to backdoor attacks has become evident. By injecting ''triggers'' to poison the training dataset, backdoor attacks manipulate the detector's prediction for inputs containing these triggers. Existing backdoor attacks against 3D object detection primarily poison 3D LiDAR signals, where large-sized 3D triggers are injected to ensure their visibility within the sparse 3D space, rendering them easy to detect and impractical in real-world scenarios. …

3d object 3d object detection abstract arxiv attacks autonomous autonomous driving backdoor become cs.cv dataset detection driving however inputs object prediction role training type vulnerability