arXiv subjects

Qiannan Wang

Publications and source records attributed to Qiannan Wang.

4 recordsLinked to original sources

arXiv2024

Path planning for autonomous vehicles with minimal collision severity

This paper proposes a path planning algorithm for autonomous vehicles, evaluating collision severity with respect to both static and dynamic obstacles. A collision severity map is generated from ratings, quantifying the severity of collisions. A two-level optimal control problem is designed. At the first level, the objective is to identify paths with the lowest collision severity. Subsequently, at the second level, among the paths with lowest collision severity, the one requiring the minimum steering effort is determined. Finally, numerical simulations were conducted using the optimal control software OCPID-DAE1. The study focuses on scenarios where collisions are unavoidable. Results demonstrate the effectiveness and significance of this approach in finding a path with minimum collision severity for autonomous vehicles. Furthermore, this paper illustrates how the ratings for collision severity influence the behaviour of the automated vehicle.

cs.RO↗

arXiv2024

SSL-OTA: Unveiling Backdoor Threats in Self-Supervised Learning for Object Detection

The extensive adoption of Self-supervised learning(SSL) has led to an increased security threat from backdoor attacks. While existing research has mainly focused on backdoor attacks in image classification, there has been limited exploration of their implications for object detection. Object detection plays a critical role in security-sensitive applications, such as autonomous driving, where backdoor attacks seriously threaten human life and property. In this work, we propose the first backdoor attack designed for object detection tasks in SSL scenarios, called Object Transform Attack (SSL-OTA). SSL-OTA employs a trigger capable of altering predictions of the target object to the desired category, encompassing two attacks: Naive Attack(NA) and Dual-Source Blending Attack (DSBA). NA conducts data poisoning during downstream fine-tuning of the object detector, while DSBA additionally injects backdoors into the pre-trained encoder. We establish appropriate metrics and conduct extensive experiments on benchmark datasets, demonstrating the effectiveness of our proposed attack and its resistance to potential defenses. Notably, both NA and DSBA achieve high attack success rates (ASR) at extremely low poisoning rates (0.5%). The results underscore the importance of considering backdoor threats in SSL-based object detection and contribute a novel perspective to the field.

cs.CR↗

arXiv2023

GhostEncoder: Stealthy Backdoor Attacks with Dynamic Triggers to Pre-trained Encoders in Self-supervised Learning

Within the realm of computer vision, self-supervised learning (SSL) pertains to training pre-trained image encoders utilizing a substantial quantity of unlabeled images. Pre-trained image encoders can serve as feature extractors, facilitating the construction of downstream classifiers for various tasks. However, the use of SSL has led to an increase in security research related to various backdoor attacks. Currently, the trigger patterns used in backdoor attacks on SSL are mostly visible or static (sample-agnostic), making backdoors less covert and significantly affecting the attack performance. In this work, we propose GhostEncoder, the first dynamic invisible backdoor attack on SSL. Unlike existing backdoor attacks on SSL, which use visible or static trigger patterns, GhostEncoder utilizes image steganography techniques to encode hidden information into benign images and generate backdoor samples. We then fine-tune the pre-trained image encoder on a manipulation dataset to inject the backdoor, enabling downstream classifiers built upon the backdoored encoder to inherit the backdoor behavior for target downstream tasks. We evaluate GhostEncoder on three downstream tasks and results demonstrate that GhostEncoder provides practical stealthiness on images and deceives the victim model with a high attack success rate without compromising its utility. Furthermore, GhostEncoder withstands state-of-the-art defenses, including STRIP, STRIP-Cl, and SSL-Cleanse.

cs.CV↗

arXiv2022

Risk-based path planning for autonomous vehicles

In this paper, a risk map-based path planning algorithm is introduced for autonomous vehicles. Multivariate B-splines are implemented to generate a risk map, which measures the risk of colliding with different objects. In the following step, a two-level optimal control problem is designed. At the first level, an overall lowest risk trajectory is found. Then in the second level, among the paths whose risk value is the lowest, the one with the minimum steering effort is determined. Finally, numerical simulations are carried out with the optimal control software OCPID-DAE1. Results show that this method is useful and meaningful to find a path of minimum risk for an autonomous vehicle.

math.OC↗