Deep Predictive Learning in Vision
Abstract
Research problem: Our existing research has identified powerful predictive learning mechanisms that enable our biologically-based co"mputational models to learn invariant object representations in a purely self-organizing manner, without requiring the massive sets" of hand-labeled images that standard deep neural networks are critically dependent upon. These mechanisms are based directly upon t"he biology of the primate visual system, and make sense of a wide range of data at many different levels of analysis. However, our e""xisting models were developed using simplified visual inputs, and we propose here to scale them up to handle large numbers of realis""tic 3D objects, ultimately with multiple objects at a time interacting with each other.Technical approach: Predictive learning requ""ires movie-like animated scenes, where the model is constantly predicting what will happen next, and learning from the difference be""tween what actually does happen and the prediction. According to our theory, this prediction-outcome learning cycle occurs every 100"" msec (10 Hz, i.e, the alpha rhythm). Thus, massive amounts of predictive learning can occur every day, providing the constraints to"" shape powerful internal generative, predictive models of the world. We propose to render large numbers of systematically generated"" scenes involving 3D objects moving with realistic physics, to train our models on increasingly complex and realistic environments."" Ultimately, our models should be able to learn from real movies and video footage, and apply their knowledge to the interpretation"" of real-world images and scenes, in a semantically rich and meaningful manner.Anticipated outcome and impacts: It is hard to overs""tate the potential impact that this project could have on DoD capabilities, by enabling multiple different mission critical technolo""gies that all depend, as people do, on the ability to effectively extract physical semantics from visual scenes. For example, this c""ould greatly facilitate automated processing of large volumes of imagery and monitoring data. Likewise, autonomous robotic systems,"" which currently depend to a large extent on human operators for their perceptual abilities, could benefit substantially. In additio""n, we anticipate learning much more about advanced visual processing in the brain.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 23, 2018
- Source ID
- N000141812116
Entities
People
- Randall O Reilly
Organizations
- Office of Naval Research
- Regents of the University of Colorado
- United States Navy