Department of Computer and Information Science
University of Pennsylvania.
We introduce a method for 'packing' salient image contours/segments
into recognizable object shapes. Most object recognition methods rely
on one-to-one matching of contours/segments to a model. However,
bottom-up image contours/segments often fragment unpredictably. We
resolve this difficulty by using many-to-one matching of image
contours to a model.
In operation, our system achieves three goals: it locates an object,
identifies its part configuration, and segments out its contours. To
learn a descriptive object shape model, we combine contours from a few
representative images. The goal is to construct a model that can be
many-to-one matched to most of the contours in the training images.
For detection, our challenges are inferring the object contours and
part locations, in addition to object location. Because the locations
of object parts and matches of contours are not annotated, they appear
as latent variables during training. We use the latent SVM learning
formulation to discriminatively tune the many-to-one matching score
using the max-margin criterion.
There are several computational implementations, using Linear
Programming (LP) or Semi-Definite Programming (SDP). We evaluate on
the challenging ETHZ shape categories dataset and outperform all
This is joint work with Qihui Zhu, Praveen Srinivasan.