Autonomous navigation systems are a testament to humanity's relentless pursuit of progress and hold immense promise. Recent developments in Deep Learning have propelled Artificial intelligence systems. However, with this promise comes a critical challenge: adversarial attacks. Autonomous navigation systems are susceptible to deliberate manipulation. These adversarial samples are data instances crafted specifically to mislead while being very similar to benign samples. Our project proposal addresses this pressing issue by exploring the adversarial robustness of autonomous navigation systems.
Indian Institute of Technology,, Mumbai – 400 076
Accessibility Controls