Kumar Krishna Agrawal
I'm a PhD student in AI at EECS, UC Berkeley, where I'm affiliated with BAIR and UCSF. I am incredibly fortunate to be advised by Prof. Adam Yala and Prof. Trevor Darrell. I also work closely with Sky Computing Lab at Berkeley, and Prof. Blake Richards at MILA.
I am interested in building efficient algorithms and systems that help improve performance of (accuracy, reliability) and access to (white-boxing, low-latency) human-centric AI. My research spans two main thrusts: 1) improving AI-assisted healthcare, with a focus on enhancing screening and diagnostic capabilities, and 2) advancing multimodal AI that integrates perception and reasoning to understand and interact with the world.
Previously, I was an AI resident at Google Brain, where I helped build fast, simple learning algorithms and systems such as data-efficient off-policy reinforcement learning for robot manipulation, deep generative models for high-fidelity audio generation, program synthesis.
Before Berkeley, I studied mathematics and computer science at IIT Kharagpur, doing reserch in multimodal deep learning, building autonomous drones and self-driving cars. I was also fortunate to spend time at MILA, advised by Prof. Yoshua Bengio.