Camilo L. Fosco

Camilo L. Fosco is a PhD candidate whose research integrates computer vision and cognitive science to build innovative models that can mimic the brain’s intricate processes to solve computer vision tasks. Currently Camilo is leveraging generative models to produce images and videos that maximize a specific cognitive metric, an approach that will enable generative techniques that are expressive and accurate, and also controllable in dimensions typically hard to quantify, like attention-grabbing potential and memorability. A MathWorks Fellowship will support his work on a ground-breaking, MATLAB-based project to analyze functional MRI and EEG data from subjects watching memorable and non-memorable short videos, which will serve as the basis for training models to regenerate the videos from the brain signals. This could lead to novel applications related to thought visualization and shed light on video processing in the brain. By combining cutting-edge generative models with insights from cognitive neuroscience, Camilo’s work has the potential to enhance our understanding of the brain’s mechanisms and help create more effective, human- centric AI solutions.