My interests broadly cover 3D computer vision and machine learning. Recently, I have focused on rethinking 3D input representations for machine learning, especially in the case of spatiotemporal data.