Shilman and Viola present an application of AdaBoost to simultaneously approach the grouping and recognition problems. The idea is simple: a group of strokes belong together if it can be recognized as something.
It's nice to see an approach that doesn't just assume the grouping problem will be solved by someone else. There are so many symbol recognition papers that make such an assumption, but not very many papers that actually provide any contribution toward solving that assumption.
When the sketches are segmented, we are happy. I agree with your discussion and I shall discuss as well. It is a less studied phenomenon, and to have the symbol grouping with any draw order is dandy.
ReplyDelete