An unsupervised neural network for low-level control of a wheeled mobile robot: noise resistance, stability, and hardware implementation

We have recently introduced a neural network mobile robot controller (NETMORC). This controller, based on previously developed neural network models of biological sensory-motor control, autonomously learns the forward and inverse odometry of a differential drive robot through an unsupervised learning-by-doing cycle. After an initial learning phase, the controller can move the robot to an arbitrary stationary or moving target while compensating for noise and other forms of disturbance, such as wheel slippage or changes in the robot's plant. In addition, the forward odometric map allows the robot to reach targets in the absence of sensory feedback. The controller is also able to adapt in response to long-term changes in the robot's plant, such as a change in the radius of the wheels. In this article we review the NETMORC architecture and describe its simplified algorithmic implementation, we present new, quantitative results on NETMORC's performance and adaptability under noise-free and noisy conditions, we compare NETMORC's performance on a trajectory-following task with the performance of an alternative controller, and we describe preliminary results on the hardware implementation of NETMORC with the mobile robot ROBUTER.

Circasemidian rather than circadian variation of circulating osteoprotegerin in clinical health

Osteoprotegerin (OPG) serves as a soluble decoy receptor for RANKL to inhibit osteoclast formation and activity. Hormones such as PTH and glucocorticoids have been reported to decrease OPG concentrations, while estrogens, transforming growth factor b, related bone morphogenic factor and thrombopoietin reportedly enhance the OPG production in the osteoblastic and bone stromal cells. Since bone turnover shows a prominent circadian rhythm in laboratory animals and humans, with bone resorption increasing at night, we investigated the time structure of circulating OPG concentrations in a group of nine healthy subjects (six women and three men; in the age range of 26-49 years). Blood samples for OPG determination were collected every 4 h for 24 h on the same day, starting at 08:00 in the morning. Data were analyzed by inferential statistical procedures, including the single and population-mean cosinor. A 12-h component was found to characterize serum OPG concentrations (P = 0.038) with peak concentrations around noon and midnight. No statistically significant circadian rhythm of OPG concentrations could be found by cosinor in our study population. The mean 24-h OPG concentration was higher in women than in men (mean +/- S.E.: 3.13 +/- 0.44 vs. 1.94 +/- 0.26 pmol/l, Student t = 2.325, P = 0.053). Since PTH concentrations also exhibit a bimodal pattern along the 24-h scale, PTH may be tested as a putative determinant of the observed changes in serum concentrations of osteoprotegerin.

Spatial asymmetries in cat retinal ganglion cell responses

Enroth-Cugell and Robson (1966) first proposed a classification of retinal ganglion cells into X cells, which exhibit approximate linear spatial summation and largely sustained responses, and Y cells, which exhibit nonlinearities and transient responses. Gaudiano (1992a, 1992b, 1994) has suggested that the dominant characteristics of both X and Y cells can be simulated with a single model simply by changing receptive field profiles to match those of the anatomical counterparts of X and Y cells. He also proposed that a significant component of the spatial nonlinearities observed in Y (and sometimes X) cells can result from photoreceptor nonlinearities coupled with push-pull bipolar connections. Specifically, an asymmetry was predicted in the ganglion cell response to rectangular gratings presented at different locations in the receptive field under two conditions: introduction/withdrawal (on-off) or contrast reversal. When measuring the response to these patterns as a function of spatial phase, the standard difference-of-Gaussians model predicts symmetrical responses about the receptive field center, while the push-pull model predicts slight but significant asymmetry in the on-off case only. To test this hypothesis, we have recorded ganglion cell responses from the optic tract fibers of anesthetized cat. The mean and standard deviations of responses to on-off and contrast-reversed patterns were compared. We found that all but one of the cells that yielded statistically significant data confirmed the hypothesis. These results largely support the theoretical prediction.

Simulations of X and Y retinal ganglion cell behavior with a nonlinear push-pull model of spatiotemporal retinal processing

This article describes a nonlinear model of neural processing in the vertebrate retina, comprising model photoreceptors, model push-pull bipolar cells, and model ganglion cells. Previous analyses and simulations have shown that with a choice of parameters that mimics beta cells, the model exhibits X-like linear spatial summation (null response to contrast-reversed gratings) in spite of photoreceptor nonlinearities; on the other hand, a choice of parameters that mimics alpha cells leads to Y-like frequency doubling. This article extends the previous work by showing that the model can replicate qualitatively many of the original findings on X and Y cells with a fixed choice of parameters. The results generally support the hypothesis that X and Y cells can be seen as functional variants of a single neural circuit. The model also suggests that both depolarizing and hyperpolarizing bipolar cells converge onto both ON and OFF ganglion cell types. The push-pull connectivity enables ganglion cells to remain sensitive to deviations about the mean output level of nonlinear photoreceptors. These and other properties of the push-pull model are discussed in the general context of retinal processing of spatiotemporal luminance patterns.