Evaluating adversarial robustness of autonomous perception systems using PyTorch, ART, and CARLA.
- Study adversarial attacks against vision models
- Evaluate robustness of object detection systems
- Integrate attacks into CARLA simulation
- Develop robustness metrics and defenses
- FGSM on MNIST
- FGSM on CIFAR-10
- PGD attacks
- Object detection attacks
- CARLA integration
- Robustness evaluation framework