Pith. sign in

REVIEW 1 cited by

Bait and Switch: Online Training Data Poisoning of Autonomous Driving Systems

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2011.04065 v2 pith:NJJZPSI5 submitted 2020-11-08 cs.LG cs.CR

Bait and Switch: Online Training Data Poisoning of Autonomous Driving Systems

classification cs.LG cs.CR
keywords onlineadversaryautonomouscausecityconceptsdatadeep
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We show that by controlling parts of a physical environment in which a pre-trained deep neural network (DNN) is being fine-tuned online, an adversary can launch subtle data poisoning attacks that degrade the performance of the system. While the attack can be applied in general to any perception task, we consider a DNN based traffic light classifier for an autonomous car that has been trained in one city and is being fine-tuned online in another city. We show that by injecting environmental perturbations that do not modify the traffic lights themselves or ground-truth labels, the adversary can cause the deep network to learn spurious concepts during the online learning phase. The attacker can leverage the introduced spurious concepts in the environment to cause the model's accuracy to degrade during operation; therefore, causing the system to malfunction.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Mirage: a Clean-Label Backdoor against LiDAR 3D Object Detection

    cs.CV 2026-06 unverdicted novelty 6.0

    Mirage achieves 73% misclassification success on LiDAR 3DOD models with 0.5% poisoning rate via label-consistent trigger injection.