论文信息 - Viewpoint Information

Viewpoint Information

In this paper, we present a new perspective to quantify the information associated with a viewpoint. The starting point is twofold: a visibility channel between a set of viewpoints and the polygons of an object, and two specific information measures introduced respectively by DeWeese and Meister (1999) and Butts (2003) to evaluate the significance of stimuli and responses in the neural code. In our approach, these information measures are applied to the visibility channel in order to quantify the information associated with each viewpoint and are compared with both viewpoint entropy and viewpoint mutual information. A number of experiments show the behavior of the proposed measures in best view selection.

Mateu Sbert | Miquel Feixas | Xavier Bonaventura

[1] M R DeWeese,et al. How to measure the information gained from one symbol. , 1999, Network.

[2] Mateu Sbert,et al. A unified information-theoretic framework for viewpoint selection and mesh saliency , 2009, TAP.

[3] Mateu Sbert,et al. Importance-Driven Focus of Attention , 2006, IEEE Transactions on Visualization and Computer Graphics.

[4] Mateu Sbert,et al. Viewpoint Selection using Viewpoint Entropy , 2001, VMV.

[5] Dimitri Plemenos,et al. Methods and data structures for virtual world exploration , 2006, The Visual Computer.

[6] M. Tarr,et al. To What Extent Do Unique Parts Influence Recognition Across Changes in Viewpoint? , 1995 .

[7] Daniel A Butts,et al. How much information is associated with a particular stimulus? , 2003, Network.

[8] Mateu Sbert,et al. Realtime automatic selection of good molecular views , 2006, Comput. Graph..

[9] M J Tarr,et al. What Object Attributes Determine Canonical Views? , 1999, Perception.