In this paper, we investigate a new approach based on WISARD Neural Network for the tracking of non-rigid deformable object. The proposed approach allows deploying an on–line training on the texture and shape features of the object, to adapt in real–time to changes, and to partially cope with occlusions. Moreover, the use of parallel classificatory trained on the same set of images allows tracking the movements of the objects. We evaluate our tracking abilities in the scenario of pizza making that represents a very challenging benchmark to test the approach since in this context the shape of the object to track completely changes during the manipulation.
Tracking deformable objects with WISARD networks
Staffa Mariacarla;
2014-01-01
Abstract
In this paper, we investigate a new approach based on WISARD Neural Network for the tracking of non-rigid deformable object. The proposed approach allows deploying an on–line training on the texture and shape features of the object, to adapt in real–time to changes, and to partially cope with occlusions. Moreover, the use of parallel classificatory trained on the same set of images allows tracking the movements of the objects. We evaluate our tracking abilities in the scenario of pizza making that represents a very challenging benchmark to test the approach since in this context the shape of the object to track completely changes during the manipulation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
