A saliency-based bottom-up visual attention model for dynamic scenes analysis
- PDF / 1,430,682 Bytes
- 20 Pages / 595.276 x 790.866 pts Page_size
- 69 Downloads / 234 Views
ORIGINAL PAPER
A saliency-based bottom-up visual attention model for dynamic scenes analysis David F. Ramirez-Moreno · Odelia Schwartz · Juan F. Ramirez-Villegas
Received: 3 October 2011 / Accepted: 13 December 2012 / Published online: 12 January 2013 © Springer-Verlag Berlin Heidelberg 2013
Abstract This work proposes a model of visual bottomup attention for dynamic scene analysis. Our work adds motion saliency calculations to a neural network model with realistic temporal dynamics [(e.g., building motion salience on top of De Brecht and Saiki Neural Networks 19:1467– 1474, (2006)]. The resulting network elicits strong transient responses to moving objects and reaches stability within a biologically plausible time interval. The responses are statistically different comparing between earlier and later motion neural activity; and between moving and non-moving objects. We demonstrate the network on a number of synthetic and real dynamical movie examples. We show that the model captures the motion saliency asymmetry phenomenon. In addition, the motion salience computation enables sudden-onset moving objects that are less salient in the static scene to rise above others. Finally, we include strong consideration for the neural latencies, the Lyapunov stability, and the neural properties being reproduced by the model.
David F. Ramirez-Moreno and Juan F. Ramirez-Villegas contributed equally to the research reported in this work. D. F. Ramirez-Moreno (B) · J. F. Ramirez-Villegas Computational Neuroscience, Department of Physics, Universidad Autonoma de Occidente, Cali, Colombia e-mail: [email protected] O. Schwartz Dominick P. Purpura Department of Neuroscience, Albert Einstein College of Medicine, Yeshiva University, New York, NY, USA O. Schwartz Department of Systems and Computational Biology, Albert Einstein College of Medicine, Yeshiva University, New York, NY, USA
Keywords Visual attention · Saliency map · Motion saliency · Neural network · Synaptic depression · Neural latency · Asymmetry phenomenon · Lyapunov stability
1 Introduction Primates’ visual cortex is capable of interpreting complex dynamical scenes in clutter. This process is thought to involve attention shifting; i.e., selecting circumscribed regions of visual information to be preferentially processed and by changing the processing focus over the time course. There have been several approaches in the literature for dynamic attention along the ventral and dorsal pathways, including both scene-dependent (bottom-up) and/or task-dependent (top-down) strategies (Itti and Koch 2000; De Brecht and Saiki 2006; Bergen and Julesz 1983; Treisman et al. 1977; Treisman and Gelade 1980); and the interactions between these two processes (Fix et al. 2010; Navalpakkam and Itti 2002, 2005, 2006; Torralba et al. 2006; Walther and Koch 2006; Wolfe et al. 2003). Many computational models of human visual search have embraced the idea of a saliency map to accomplish preattentive selection. This representation contains the overall neural activity elicited by objects and no
Data Loading...
