Work

TrojDRL: Evaluation of Backdoor Attacks on Deep Reinforcement Learning

Year: 2020
Type: conference-paper
Abstract: We present TrojDRL, a tool for exploring and evaluating backdoor attacks on deep reinforcement learning agents. TrojDRL exploits the sequential nature of deep reinforcement learning (DRL) and considers different gradations of threat models. We show that untargeted attacks on state-of-the-art actor-critic algorithms can circumvent existing defenses built on the assumption of backdoors being targeted. We evaluated TrojDRL on a broad set of DRL benchmarks and showed that the attacks require only poisoning as little as 0.025% of the training data. Compared with existing works of backdoor attacks on classification models, TrojDRL provides a first step towards understanding the vulnerability of DRL agents.
Language: English
Open Access status: green
OpenAlex ID: W3092005433
FWCI: 3.456
Cites: 25
Cited by: 137
Related to: 10