SA-Attack: Speed-adaptive stealthy adversarial attack on trajectory prediction
This addresses a safety-critical problem for automated vehicle navigation by introducing a more realistic and stealthy adversarial attack, though it is incremental as it builds on prior attack methods by adding speed adaptation and stealth features.
The paper tackles the vulnerability of neural network-based trajectory prediction models in automated vehicles by proposing SA-Attack, a speed-adaptive stealthy adversarial attack method that achieves high attack success rates while ensuring adaptability to realistic speed scenarios and concealment of the attacks.
Trajectory prediction is critical for the safe planning and navigation of automated vehicles. The trajectory prediction models based on the neural networks are vulnerable to adversarial attacks. Previous attack methods have achieved high attack success rates but overlook the adaptability to realistic scenarios and the concealment of the deceits. To address this problem, we propose a speed-adaptive stealthy adversarial attack method named SA-Attack. This method searches the sensitive region of trajectory prediction models and generates the adversarial trajectories by using the vehicle-following method and incorporating information about forthcoming trajectories. Our method has the ability to adapt to different speed scenarios by reconstructing the trajectory from scratch. Fusing future trajectory trends and curvature constraints can guarantee the smoothness of adversarial trajectories, further ensuring the stealthiness of attacks. The empirical study on the datasets of nuScenes and Apolloscape demonstrates the attack performance of our proposed method. Finally, we also demonstrate the adaptability and stealthiness of SA-Attack for different speed scenarios. Our code is available at the repository: https://github.com/eclipse-bot/SA-Attack.