Single Photon Avalanche Diode: An Avalanche of Photon Data for Sharper Imaging

While CMOS sensors have been the industry norm for camera systems, single photon avalanche diode (SPAD) sensors are gaining traction as the new gold standard. Sensor design is advancing toward higher resolution and more integrated camera architectures, as demonstrated by new SPAD efforts from companies such as Canon and Pi Imaging Technology (a ZEISS company).

However, these new sensors collect a significant amount of data. For instance, a one-megapixel sensor can capture one million photons per second, creating a terabyte of data per second. This session will explore how SPADs work, how their performance compares to CMOS sensors, and how SPAD sensor data can be transformed into perception outputs for sharper imaging that advances computer vision.

About the presenter
Sebastian Bauer, Ph.D., is the co-founder & CEO of Ubicept, a startup transforming raw sensor data into high-quality perception outputs that keep pace with rapidly advancing AI systems. With a background in engineering, Bauer has always thrived at the intersection of cutting-edge research and real-world innovation, transforming groundbreaking ideas into tangible technology.

In 2018, he joined the University of Wisconsin–Madison (UW) as a postdoc, focusing on the processing of photon data. Three years later, Bauer teamed up with UW professors Mohit Gupta and Andreas Velten, MIT professor Ramesh Raskar, and MIT doctoral graduate Tristan Swedish to launch Ubicept. The company’s mission is to give the world access to the most advanced optical perception physics allows.