Neural Networks for Model-Based Recognition
Abstract
This annual progress report describes first-year progress in theoretical and applied fronts for neutral-net object recognition via graph matching. On the theory front, a learning scheme is applied to our previously hand-designed graphs, and a Bayesian approach to weighted graph matching is described. On an applied front, our networks are applied to recognition of machined parts. Continuing progress on the application of continuation optimization methods to our networks is reported.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 12, 1991
- Accession Number
- ADA241900
Entities
People
- Gene R. Gindi
Organizations
- Yale University