Cedric Scheerlinck

I am a deep learning research engineer at Luma AI training generative multimodal foundation models capable of reasoning across vision and language.

Previously I worked at Zoox on AI models that enable robotaxi to predict the intention and future behavior of surrounding vehicles, pedestrians etc. to drive safely and fully autonomously.

Before that I was a senior deep learning engineer at Skydio for three years working on autonomous drones for inspection, monitoring, search and rescue. I worked on the full data/training/evaluation pipeline for perception, semantic segmentation, 3d reconstruction and obstacle avoidance. I trained a novel night obstacle avoidance model that enables the drone to see in the dark, leading to the successful launch of a new product feature: NightSense.

I completed my PhD at the Australian National University and the Australian Center for Robotic Vision in 2021. My research was on Robotic Vision: image processing to allow robots to see and make sense of the world. I created novel algorithms and AI models to decode visual signals from event cameras (novel sensors that work like a human eye), ultimately unlocking high-speed, HDR, real-time video/perception from tiny efficient sensors.