Probabilistic Solution of Ill-Posed Problems in Computational Vision.
Abstract
Computational vision is a set of inverse problems. The authors review standard regularization theory, discuss its limitations, and present new stochastic (in particular, Bayesian) methods for their solution. They derive efficient algorithms and describe parallel implementations on digital parallel SIMD architectures, as well as a new class of parallel hybrid computers. Keywords: Stochastic methods; Artificial intelligence; Problem solving; Probablistic approach. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1987
- Accession Number
- ADA183807
Entities
People
- J. Marroquin
- Sanjoy K. Mitter
- Tomaso Poggio
Organizations
- Massachusetts Institute of Technology