Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems, 2010
Does the dynamical regime in which a system engages when it is coping with a situation A change a... more Does the dynamical regime in which a system engages when it is coping with a situation A change after adaptation to a new situation B? Is homeostatic instability a generic mechanism for flexible switching between dynamical regimes? We develop a model to approach these questions where a simulated agent that is stable and performing phototaxis has its vision field inverted so that it becomes unstable; instability activates synaptic plasticity changing the agent’s simulated nervous system attractor landscape towards a configuration that accommodates stable dynamics under normal and inverted vision. Our results show that: 1) the dynamical regime in which the agent engages under normal vision changes after adaptation to inverted vision; 2) homeostatic instability is not necessary for switching between dynamical regimes. Additionally, during the dynamical system analyses we also show that: 3) qualitatively similar behaviours (phototaxis) can be generated by different dynamics; 4) the agent’s simulated nervous system operates in transient dynamic towards an attractor that continuously move on the phase space; and 5) plasticity moves and reshapes the attractor landscape in order to accommodate a stable dynamical regimes to deal with inverted vision.
... Lincoln Smith, Andrew Philippides, Paul Graham, and Phil Husbands ... Biological Cybernetics ... more ... Lincoln Smith, Andrew Philippides, Paul Graham, and Phil Husbands ... Biological Cybernetics 95(5), 413–430 (2006) 7. Lambrinos, D., Möller, R., Pfeifer, R., Wehner, R., Labhart, T.: A mobile robot employing insect strategies for navigation. ...
To behave in a robust and adaptive way, animals must extract task-relevant sensory information ef... more To behave in a robust and adaptive way, animals must extract task-relevant sensory information efficiently. One way to understand how they achieve this is to explore regularities within the information animals perceive during natural behavior. In this chapter, we describe how we have used artificial neural networks (ANNs) to explore efficiencies in vision and memory that might underpin visually guided route navigation in complex worlds. Specifically, we use three types of neural network to learn the regularities within a series of views encountered during a single route traversal (the training route), in such a way that the networks output the familiarity of novel views presented to them. The problem of navigation is then reframed in terms of a search for familiar views, that is, views similar to those associated with the route. This approach has two major benefits. First, the ANN provides a compact holistic representation of the data and is thus an efficient way to encode a large set of views. Second, as we do not store the training views, we are not limited in the number of training views we use and the agent does not need to decide which views to learn.
We are working on the ESA Bionics and Space System Design contract AO/1-4469/03/NL/SFe . The proj... more We are working on the ESA Bionics and Space System Design contract AO/1-4469/03/NL/SFe . The project is to review the application of biomimetics to space missions. Our particular focus is on robotic activities with an emphasis on planetary exploration. At the time of ASTRA ...
The authors are particularly grateful to the surviving members of the Ratio Club, Horace Barlow, ... more The authors are particularly grateful to the surviving members of the Ratio Club, Horace Barlow, John Westcott, and Philip Woodward, who generously participated in the research for this paper, and to the late Jack Good, Harold Shipton, and Tommy Gold, all of whom we ...
Artificial Life XII: Proceedings of the Twelfth International Conference on the Synthesis and Simulation of Living Systems, 2010
Does the dynamical regime in which a system engages when it is coping with a situation A change a... more Does the dynamical regime in which a system engages when it is coping with a situation A change after adaptation to a new situation B? Is homeostatic instability a generic mechanism for flexible switching between dynamical regimes? We develop a model to approach these questions where a simulated agent that is stable and performing phototaxis has its vision field inverted so that it becomes unstable; instability activates synaptic plasticity changing the agent’s simulated nervous system attractor landscape towards a configuration that accommodates stable dynamics under normal and inverted vision. Our results show that: 1) the dynamical regime in which the agent engages under normal vision changes after adaptation to inverted vision; 2) homeostatic instability is not necessary for switching between dynamical regimes. Additionally, during the dynamical system analyses we also show that: 3) qualitatively similar behaviours (phototaxis) can be generated by different dynamics; 4) the agent’s simulated nervous system operates in transient dynamic towards an attractor that continuously move on the phase space; and 5) plasticity moves and reshapes the attractor landscape in order to accommodate a stable dynamical regimes to deal with inverted vision.
... Lincoln Smith, Andrew Philippides, Paul Graham, and Phil Husbands ... Biological Cybernetics ... more ... Lincoln Smith, Andrew Philippides, Paul Graham, and Phil Husbands ... Biological Cybernetics 95(5), 413–430 (2006) 7. Lambrinos, D., Möller, R., Pfeifer, R., Wehner, R., Labhart, T.: A mobile robot employing insect strategies for navigation. ...
To behave in a robust and adaptive way, animals must extract task-relevant sensory information ef... more To behave in a robust and adaptive way, animals must extract task-relevant sensory information efficiently. One way to understand how they achieve this is to explore regularities within the information animals perceive during natural behavior. In this chapter, we describe how we have used artificial neural networks (ANNs) to explore efficiencies in vision and memory that might underpin visually guided route navigation in complex worlds. Specifically, we use three types of neural network to learn the regularities within a series of views encountered during a single route traversal (the training route), in such a way that the networks output the familiarity of novel views presented to them. The problem of navigation is then reframed in terms of a search for familiar views, that is, views similar to those associated with the route. This approach has two major benefits. First, the ANN provides a compact holistic representation of the data and is thus an efficient way to encode a large set of views. Second, as we do not store the training views, we are not limited in the number of training views we use and the agent does not need to decide which views to learn.
We are working on the ESA Bionics and Space System Design contract AO/1-4469/03/NL/SFe . The proj... more We are working on the ESA Bionics and Space System Design contract AO/1-4469/03/NL/SFe . The project is to review the application of biomimetics to space missions. Our particular focus is on robotic activities with an emphasis on planetary exploration. At the time of ASTRA ...
The authors are particularly grateful to the surviving members of the Ratio Club, Horace Barlow, ... more The authors are particularly grateful to the surviving members of the Ratio Club, Horace Barlow, John Westcott, and Philip Woodward, who generously participated in the research for this paper, and to the late Jack Good, Harold Shipton, and Tommy Gold, all of whom we ...
This paper describes investigations into using evolutionary search for quantitative spectroscopy.... more This paper describes investigations into using evolutionary search for quantitative spectroscopy. Given the spectrum (intensity × frequency) of a sample of material of interest, we would like to be able to infer the make-up of the material in terms of percentages by mass of its constituent compounds. The problem is usually tackled using regression methods. This approach can have various
Uploads
Chapters by Phil Husbands
Papers by Phil Husbands