Case Study
Tuesday, July 01
08:30 AM - 09:00 AM
Live in San Francisco
Less Details
Radars are crucial for achieving safe and reliable autonomous driving because of their high sensitivity to detecting moving objects, whether occluded or at long ranges, and their robustness in adverse weather conditions. However, radars also face challenges such as low spatial resolution and high noise levels in the data, which limits the performance of radar-based perception AI models. So, how can we best utilize radars in autonomous vehicles?
This talk explores the current state-of-the-art in radar-based AI for autonomy and highlights promising new directions that could transform the use of radars in autonomous vehicles.
I lead research and development of AI perception models for autonomous vehicles. I apply my expertise in AI to develop novel transformers cross-attention based early sensor fusion models.
During my masters at Stanford, I focused on radar perception methods and AI models development. Before joining Zoox, I co-founded Werewolf AI, a B2B startup that created realistic rare-class datasets for autonomous companies. I am passionate about innovating and advocating for a safer and more sustainable future of our cities supported by autonomous vehicles.