C. Scheier and D. Lambrinos. Categorization in a real-world agent using haptic exploration and active perception. In Proc. of the 4th Int. Conf. on Simulation of Adaptive Behavior (SAB'96), pages 65--75. Cambridge, MA: MIT Press, 1996.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... perceptual categorization MR AG Sporns et al. 30] neuromodulation MR AG Sporns and Alexander [31] Categorization sensorimotor categorization AVH Berthouze and Kuniyoshi [32] invariant object recognition MR AG Krichmar and Edelman [28] sensorimotor categorization MR AG Scheier and Lambrinos [33] developmental plasticity role of behavioral interaction MR AG Almassy et al. 34] obstacle avoidance, sensory deprivation MR Elliott and Shadbolt [35] perceptual categorization MR AG Sporns et al. 30] Motor skill acquisition pendulation, morphological changes HD Lungarella and Berthouze ....

....[see 47] In this sense, this area of research represents a subset of the one related to sensorimotor control. Categorization of objects via real time correlation of temporally contingent information impinging on the robot s haptic and visual sensors was achieved by Scheier and Lambrinos [33]. The suggested robot control architecture employed sensorimotor coordination at various functional levels for saccading on interesting regions in the environment, for attentional sensorimotor loops, and for category learning. Sensorimotor activity was also critical in work performed by ....

[Article contains additional citation context not shown here]

C. Scheier and D. Lambrinos. Categorization in a real-world agent using haptic exploration and active perception. In Proc. of the 4th Int. Conf. on Simulation of Adaptive Behavior (SAB'96), pages 65--75. Cambridge, MA: MIT Press, 1996.

....is not only not dependent upon internal arbitration, but also not dependent on any kind of internalization of behaviour at all . The argument this time is how such a technique could work in a real environment, given that the sensor motor processes are not hand designed. Scheier and Lambrinos [30] proposed quite a similar perspective, which works in a real environment. Objects are categorized through exploration, even though each individual process still has to be explicitly defined, as well as the various combinations of processes delivering coherent behaviour. Seth argues that his ....

....a huge search space of possible configurations, though even in this case there is still a need to obtain a formal representation of actions and their rewarding constants through experimentation. 4. 3 Behaviour Scheme The models described are all implemented in artificial environments (except [30]) providing suggestions and inspiration to the present study. The behaviour scheme decided for the model is based on the bottom up approach hierarchy as described by Maes [39] but without any internal representation of behaviours as in [30,41] while paying more attention to ethological concerns ....

[Article contains additional citation context not shown here]

Scheier, C., Lambrinos, D. "Categorization in a Real-World Agent Using Haptic Exploration and Active Vision", Proceedings of the Fourth International Conference on Simulation of Adaptive Behaviour, pp 65-74,

....and selectively decide when to use feedback to revise its sensorimotor mapping. According to the author, this mechanism allows the animat to exibly assign varying meanings to environmental stimuli. Finally, two speci c aspects of the perception action coordination problem are dealt with in SCHE96 and WILS00. Scheier and Lambrinos describe how categorization must be treated as a sensori motor coordination problem, not as a perceptual one. Their system learns to recognize an object with visual and haptic modalities and to discriminate it from other objects (SCHE96) Wilson and Neal, using a ....

....problem are dealt with in SCHE96 and WILS00. Scheier and Lambrinos describe how categorization must be treated as a sensori motor coordination problem, not as a perceptual one. Their system learns to recognize an object with visual and haptic modalities and to discriminate it from other objects (SCHE96) Wilson and Neal, using a model of interactions between a shepherd, his dog and a sheep, study how the behavior repertoire of the dog robot impacts on the number of interactions required from the shepherd to control the sheep (WILS00) 3. Action selection and behavioral sequences Several ....

C. Scheier and D. Lambrinos. Categorization in a Real-World Agent Using Haptic Exploration and Active Perception. In [SAB96].

....since through its own movement, the agent generates correlations in the interaction with the environment. The second important aspect of sensorymotor coordination is the generation of cross modal associations, including proprioceptive cues originating from the motor system (Thelen and Smith, 1994; Scheier and Lambrinos, 1996). Figure 9: Infant categorizing objects and building up concepts while engaged in sensory motor coordination. Additional support for the principle of sensory motor coordination comes from developmental studies. There is a lot of evidence that concept formation in human infants is directly based on ....

....in introspection. The latter has long ago been demonstrated to be a poor guide to research (Nisbett and Wilson, 1977) This principle is supported by numerous research contributions (e.g. Ballard, 1991; Dewey, 1896; Douglas, 1993; Edelman, 1987; Thelen and Smith, 1994; Smith and Thelen, 1993; Scheier and Lambrinos, 1996; Pfeifer and Scheier, in press; Scheier and Pfeifer, 1995) The principle of ecological balance The principle of ecological balance states that there has to be a match between the complexity of the sensors, the actuators, and the neural substrate. Moreover, it states that the tasks have to ....

Scheier, C., and Lambrinos, D. (1996). Categorization in a real-world agent using haptic exploration and active vision. Proc. SAB'96.

....are to be adaptive, their complexity will have to be increased. We have applied a similar kind of analysis to a more complex robot with a camera and more complex motor control (the SMC II bot ) using a value based learning scheme, similar to the ones use in the Darwin 3 system [8] see [9]) There, crossmodal learning took place between the visual and the haptic modality. We have shown that up to a certain degree, through learning, the co activation in the haptic and visual feature maps is increased and thus the redundancy always increases up to a certain value. This implies that ....

Scheier, C., and Lambrinos, D. (1996). Categorization in a real-world agent using haptic exploration and active perception. In Proc. SAB 96, 65-74.

....to these problems, several theoretical problems will be addressed in SMC II which have not been dealt with in SMC I. SMC II: Learned reentrant mappings and attentional sensorymotor loops Summary The SMC II is an extention of the SMC I models. It is described in detail in Scheier and Lambrinos ([27]) Here we give a summary of the main concepts and results. In essence, the improvements consist in an increase in the sophistication of the potential sensory motor coordinations. First, the sensory motor complexity of the robot is significantly augmented. While the sensory system in the previous ....

....evolved it still explored some of the objects. After having encountered around 12 non conductive objects the robot only explored because of errors in the sensory readings. A detailed analysis of the internal dynamics underlying this learning process can be found in Scheier and Lambrinos ([27]) 20 Discussion We now relate SMC II back to the functions of sensory motor coordination outlined in the introduction. For reasons of space we only discuss three of them. First, a foveation mechanism is used to focus the fovea on interesting regions in the environment. The sensory motor ....

Scheier, C. and Lambrinos, D. (1996). Categorization in a real-world agent using haptic exploration and active perception. In: Proc. SAB96.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC