Object tracking using learned feature manifolds

• A local feature based manifold representation for object tracking.
• Learn a manifold for each feature and approximate it by a set of linear subspaces.
• Represent the object by a manifold graph which encodes object structures.
• Apply manifold representation to tracking and update manifold status adaptively.
• Show our tracking results and compare our tracker with state-of-the-art methods.

Local feature based object tracking approaches have been promising in solving the tracking problems such as occlusions and illumination variations. However, existing approaches typically model feature variations using prototypes, and this discrete representation cannot capture the gradual changing property of local appearance. In this paper, we propose to model each local feature as a feature manifold to characterize the smooth changing behavior of the feature descriptor. The manifold is constructed from a series of transformed images simulating possible variations of the feature being tracked. We propose to build a collection of linear subspaces which approximate the original manifold as a low dimensional representation. This representation is used for object tracking. Object location is located by a feature-to-manifold matching process. Our tracking method can update the manifold status, add new feature manifolds and remove expiring ones adaptively according to object appearance. We show both qualitatively and quantitatively this representation significantly improves the tracking performance under occlusions and appearance variations using standard tracking dataset.