Contextual Recognition of Hand-Drawn Diagrams with Conditional Random Fields

Authors:
Martin Szummer;Yuan Qi
Affiliations:
Microsoft Research;Microsoft Research
Venue:
IWFHR '04 Proceedings of the Ninth International Workshop on Frontiers in Handwriting Recognition
Year:
2004

Citing 0
Cited 8

Learning Diagram Parts with Hidden Random Fields

ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
Handwritten Gesture Recognition Driven by the Spatial Context of Strokes

ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
Discriminative Random Fields

International Journal of Computer Vision
Bibliography

ACM SIGGRAPH 2007 courses
Modelling sequences using pairwise relational features

Pattern Recognition
Technical Section: Using data mining for digital ink recognition: Dividing text and shapes in sketched diagrams

Computers and Graphics
Semi-supervised learning for image annotation based on conditional random fields

CIVR'06 Proceedings of the 5th international conference on Image and Video Retrieval
Sketched symbol recognition using Latent-Dynamic Conditional Random Fields and distance-based clustering

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

Hand-drawn diagrams present a complex recognition problem. Fragments of the drawing are often individually ambiguous, and require context to be interpreted. We present a recognizer based on conditional random fields (CRFs) that jointly analyze all drawing fragments in order to incorporate contextual cues. The classification of each fragment influences the classification of its neighbors. CRFs allow flexible and correlated features, and take temporal information into account. Training is done via conditional MAP estimation that is guaranteed to reach the global optimum. During recognition we propagate information globally to find the joint MAP or maximum marginal solution for each fragment. We demonstrate the framework on a container versus connector recognition task.