Towards Contextual Action Recognition and Target Localization with Active Allocation of Attention
Exploratory gaze movements are fundamental for gathering the most relevant information regarding the partner during social interactions. We have designed and implemented a system for dynamic attention allocation which is able to actively control gaze movements during a visual action recognition task. During the observation of a partner’s reaching movement, the robot is able to contextually estimate the goal position of the partner hand and the location in space of the candidate targets, while moving its gaze around with the purpose of optimizing the gathering of information relevant for the task. Experimental results on a simulated environment show that active gaze control provides a relevant advantage with respect to typical passive observation, both in term of estimation precision and of time required for action recognition.
Keywordsactive vision social interaction humanoid robots attentive systems information gain
Unable to display preview. Download preview PDF.
- 2.Ballard, D.H.: Animate vision. AI 48, 57–86 (1991)Google Scholar
- 9.Kwok, C., Fox, D.: Reinforcement learning for sensing strategies. In: IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2004 (2004)Google Scholar
- 11.Marr, D.: Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. W. H. Freeman, New York (1982)Google Scholar
- 12.Ognibene, D., Balkenius, C., Baldassarre, G.: Integrating epistemic action (active vision) and pragmatic action (reaching): A neural architecture for camera-arm robots. In: Proceedings of the Tenth International Conference on the Simulation of Adaptive Behavior (2008)Google Scholar
- 13.Ognibene, D., Wu, Y., Lee, K., Demiris, Y.: Hierarchies for embodied action perception. Under review (2012)Google Scholar
- 14.Ognibene, D., Pezzulo, G., Baldassarre, G.: How can bottom-up information shape learning of top-down attention control skills? In: Proceedings of 9th International Conference on Development and Learning (2010)Google Scholar
- 16.Sarabia, M., Ros, R., Demiris, Y.: Towards an open-source social middleware for humanoid robots. In: Proc. 11th IEEE-RAS Int Humanoid Robots (Humanoids) Conf., pp. 670–675 (2011)Google Scholar
- 17.Schmidhuber, J., Huber, R.: Learning to generate artificial fovea trajectories for target detection. Int. J. Neural Syst. 2(1-2), 135–141 (1991)Google Scholar
- 18.Sommerlade, E., Reid, I.: Information theoretic active scene exploration. In: Proc. IEEE Computer Vision and Pattern Recognition (CVPR) (May 2008)Google Scholar