No abstract

Deep reinforcement learning methods have been widely used in recent years for autonomous vehicle's decisionmaking. A key issue is that deep neural networks can be fragile to adversarial attacks or other unseen inputs. In this paper, we address the latter issue: we focus on generating socially acceptable perturbations (SAP), so that the autonomous vehicle (AV agent), instead of the challenging vehicle (attacker), is primarily responsible for the crash. In our process, one attacker is added to the environment and trained by deep reinforcement learning to generate the desired perturbation. The reward is designed so that the attacker aims to fail the AV agent in a socially acceptable way. After training the attacker, the agent policy is evaluated in both the original naturalistic environment and the environment with one attacker. The results show that the agent policy which is safe in the naturalistic environment has many crashes in the perturbed environment.

A highly automated vehicle (HAV) is a safety-critical system. Therefore, a verification and validation (V&V) process that rigorously evaluates the safety of HAVs is necessary before their release to the market. In this paper, we propose an interaction-aware safety evaluation framework for the HAV and apply it to the roundabout entering, a highly interactive driving scenario with various traffic situations. Instead of assuming that the primary other vehicles (POVs) take predetermined maneuvers, we model the POVs as game-theoretic agents. To capture a wide variety of interactions between the POVs and the vehicle under test (VUT), we use level-k game theory and social value orientation to characterize the interactive behaviors and train a diverse library of POVs using reinforcement learning. The game-theoretic library, together with initial conditions, form a rich testing space for the two-POV roundabout scenario. On the other hand, we propose an adaptive test case generation scheme based on adaptive sampling and stochastic optimization to efficiently generate customized challenging cases for the VUT from the testing space. In simulations, the proposed testing space design captured a wide range of interactive situations at the roundabout scenario. The proposed test case generation scheme was found to cover the failure modes of the VUT more effectively compared to other test case generation approaches.

This study aimed to investigate the effect of EBI3 on radiation-induced immunosuppression of cervical cancer HeLa cells by regulating Treg cells through PD-1/PD-L1 signaling pathway. A total of 43 adult female Wistar rats were selected and injected with HeLa cells in the caudal vein to construct a rat model of cervical cancer. All model rats were randomly divided into the radiotherapy group ( n = 31) and the control group ( n = 12). The immunophenotype of Treg cells was detected by the flow cytometry. The protein expressions of EBI3, PD-1, and PD-L1 in cervical cancer tissues were tested by the streptavidin-peroxidase method. HeLa cells in the logarithmic growth phase were divided into four groups: the blank, the negative control group, the EBI3 mimics group, and the EBI3 inhibitors group. Western blotting was used to detect PD-1 and PD-L1 protein expressions. MTT assay was performed to measure the proliferation of Treg cells. Flow cytometry was used to detect cell cycle and apoptosis, and CD4/CD8 T cell ratio in each group. Compared with before and 1 week after radiotherapy, the percentages of CD4T cells and CD8T cells were significantly decreased in the radiotherapy group at 1 month after radiotherapy. Furthermore, down-regulation of EBI3 and up-regulation of PD-1 and PD-L1 were observed in cervical cancer tissues at 1 month after radiotherapy. In comparison to the blank and negative control groups, increased expression of EBI3 and decreased expressions of PD-1 and PD-L1 were found in the EBI3 mimics group. However, the EBI3 inhibitors group had a lower expression of EBI3 and higher expressions of PD-1 and PD-L1 than those in the blank and negative control groups. The EBI3 mimics group showed an increase in the optical density value (0.43 ± 0.05), while a decrease in the optical density value (0.31 ± 0.02) was found in the EBI3 inhibitors group. Moreover, compared with the blank and negative control groups, the apoptosis rates of Treg/CD4T/CD8T cells were decreased in the EBI3 mimics group, but the EBI3 inhibitors group exhibited an increase in apoptosis rate. In conclusion, over-expression of EBI3 could reduce the apoptosis of Treg/CD4T/CD8T cells and prevent radiation-induced immunosuppression of cervical cancer HeLa cells by inhibiting the activation of PD-1/PD-L1 signaling pathway.

No abstract