Think like an amateur. Do as an expert. – Takeo Kanade
I got a chance to see Takeo Kanade talk today at UCF. He was fantastic! Takeo Kanade is one of the highest ranked/cited computer scientists on computer vision and robotics. He teaches at Carnegie Mellon University. He talked about various projects he worked on over the years, and told a few inspiring stories.
Although many people believe that he got his ideas for the EyeVision project from the Wachowskis’ cinematography in The Matrix (because that is how it was described during the Super Bowl), it is a fact that Kanade and his students had already published similar many-camera systems as early as 1995. Here is a video showing the EyeVision project used during Super Bowl XXXV:
Kanade also told an inspiring story about how his persistent graduate student Bruce Lucas kept pushing an optimization that he had made to an early computer vision algorithm circa 1980, but because Kanade “was such an expert at the time” he dismissed his optimizations at the time, finding them to be trivial. Eventually he conceeded to let Lucas publish with his name, and the result historically is that the Lucas-Kanade method is one of the most cited computer vision algorithms in the world (well over 10000 citations on Google Scholar). This story is similar to David Huffman’s contributions and Huffman compression.
Lucas, B. D., & Kanade, T. (1981). An iterative image registration technique with an application to stereo vision.
Here are various other videos about Kanade-san: