Aude Oliva

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Aude Oliva, Ph.D.
Known forcomputer vision, hybrid images
Scientific career
ThesisPerception de scènes : traitement fréquentiel du signal visuel : aspects psychophysiques et neurophysiologiques
Doctoral advisorJeanny Herault

Aude Oliva is a French professor of computer vision, neuroscience, and human-computer interaction at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL).


Oliva has a dual French baccalaureate in mathematics and physics. She then earned a Masters of Science in experimental psychology and cognitive neuroscience from the Institut National Polytechnique in Grenoble and then a doctorate from the same university in 1995.[1] She joined the MIT faculty in 2004 and CSAIL in 2012.[2]


Oliva's group investigates how psychological perception of an image can change based on memorability, content, and limitations of human visual systems. Her most recognizable work is the hybrid image, the classic example of which combines the high-frequency outline and detail of Albert Einstein's face with a blurry, low-frequency image of Marilyn Monroe; the latter becomes focused only when viewed from long distances. Such images have found use in information privacy, time-lapses, marketing, and brainteasers.[3] Another branch of her research deals with object-vs-scene image processing in human brains, where Oliva and others have postulated that part of our visual system focuses on a fast recognition and classification of a familiar scene (birthday party) rather than individual component objects in the scene (cake).[4]

She has most recently used deep learning to teach computers how to recognize locations[5] in an image by a combination of its features. For example, a bed, window, and posters might indicate a bedroom, while a stove, tile, and countertop might indicate a kitchen.[citation needed]

Oliva's work has been used by those in the field of artificial imagination, the concept of building a human-like consciousness from computer algorithms.[citation needed]



  1. ^ Brahim, Naouraz; Daniel, Sylvie; Guériot, Didier; Solaiman, Basel (2012-12-28). "Vers une méthodologie de reconstruction 3D de scènes sous-marines par caméras. Problématiques et études préliminaires". Traitement du Signal. 29 (6): 521–551. doi:10.3166/ts.29.521-551. ISSN 0765-0019.
  2. ^ a b "John Simon Guggenheim Foundation Aude Oliva". Archived from the original on 2018-11-11. Retrieved 2018-11-10.
  3. ^ Oliva, Aude (2006). "Hybrid Images" (PDF). MIT CSAIL. Archived (PDF) from the original on 2018-03-04.
  4. ^ Greene, Michelle R.; Oliva, Aude (March 2009). "Recognition of natural scenes from global properties: Seeing the forest without representing the trees". Cognitive Psychology. 58 (2): 137–176. doi:10.1016/j.cogpsych.2008.06.001. ISSN 0010-0285. PMC 2759758. PMID 18762289.
  5. ^ "Dissecting Artificial Intelligence to Better Understand the Human Brain - Cognitive Neuroscience Society". Cognitive Neuroscience Society. 2018-03-25. Archived from the original on 2018-03-31. Retrieved 2018-11-10.
  6. ^ "2016 Vannevar Bush Fellows". Retrieved 2018-11-10.
  7. ^ "Aude Oliva | Simons Institute for the Theory of Computing". Retrieved 2018-11-10.