Rate My Professor Pietro Perona

Directing the Perona Lab, his research delves into the computational foundations of vision, creating machine vision systems for visual recognition and categorization. These algorithms allow machines to learn to identify objects like frogs, cars, faces, and trees from minimal examples and human guidance. The Visipedia project has yielded widely used apps iNaturalist and Merlin Bird ID for species recognition from photographs. In collaborations, he develops tools to analyze behaviors of fruit flies and mice, connecting genes, brains, and actions. Perona studies human visual performance in search and recognition tasks, including web-based crowdsourcing for image annotation. Committed to responsible AI, he devises methods to evaluate accuracy and bias in face recognition and computer vision. He teaches graduate courses including Deep Learning (EE/CNS/CS 148) and Vision: From Computational Theory to Neuronal Mechanisms (CNS/Bi/EE/CS/NB 186). His impactful publications encompass Microsoft COCO: Common Objects in Context (2014, 70,763 citations), Scale-space and edge detection using anisotropic diffusion (2002, 18,198 citations), The Caltech-UCSD Birds-200-2011 Dataset (2011, 5,817 citations), Learning generative visual models from few training examples (2004), and A Bayesian hierarchical model for learning natural scene categories (2005). Honors include IEEE Fellow (2020), PAMI Distinguished Researcher Award (2021), Best Paper Award (2019), honorary degree (2025), and Stibitz-Wilson Award (2025).