home › event - a graph lattice approach to maintaining dense collections of subgraphs as image features

EVENT:

A graph lattice approach to maintaining dense collections of subgraphs as image features
Conferences & Talks

 

description

Document classification and indexing methods depend on having informative image features. This paper shows how large families of complex features can be built out of simpler ones through construction of a graph lattice -- a hierarchy of related subgraphs linked in a lattice. A graph lattice enables efficiency gains that make it possible to effectively employ bag-of-words methods for document classification using high-dimensional feature vectors. Each feature is itself a subgraph, and a feature vector is a count of occurrences of subgraphs in the image. The graph lattice enables methods for adaptively growing a feature space of subgraphs tailored to observed document genres. We demonstrate the approach through classification of forms containing rectilinear line art.

 

upcoming events   view all 

Making Robots Work to Help us Work Remotely
Leila Takayama, Susan Herring, Dallas Goecker, Victoria Bellotti
19 October 2017 | George E. Pake Auditorium, PARC
PARC Forum  

Innovation and AI
Tolga Kurtoglu
5 November 2017 | Lisbon, Portugal
Conferences & Talks  

Leveraging RF power for flexible-hybrid electronics
6 November 2017
Conferences & Talks  

The Future of Electronics
Mike Kuniavsky, Janos Veres
14 November 2017 | San Francisco, CA
Conferences & Talks  

Printed Electronics USA 2017 - Visit PARC's Booth #X22
Ross Bringans, Markus Larsson, Nicholas Meehan, Janos Veres
15 November 2017 - 16 November 2017 | Santa Clara, CA
Conferences & Talks