Artificial neural networks for coordination and control: The portability of experiential representations

It is time to locate connectionist representation theory in the new wave of robotics research. The utility of representations developed in artificial neural networks (ANNs) during learning has been demonstrated in cognitive science research since the 1980s. The research reported here puts learned representations to work in a decentered control task, the disembodied arm problem, in which a mobile robot operates an arm fixed to a table to pick up objects. There is no physical linkage between the arm and the robot and so the robot's point of view must be decentered. This is done by developing a modular Artificial Neural Net system in three stages: (i) a classifier net is trained with laser scan data to output transformationally invariant position classes; (ii) an arm net is trained for picking up objects; (iii) an inter net is trained to communicate and coordinate the sensing and acting. The completed system is shown to create new nonsymbolic transformationally invariant representations in order to perform the effective generalization of decentered viewpoints.     

ER--Regina style. Interview by Pamela Cowan

Two hundred and fifteen patients with pulmonary involvement of lobar (regional, segmental) extent were examined. The examination procedure included clinical and laboratory tests, X-ray tomographic study, and computed tomography, bronchoscopy with biopsy, cyto- and histological studies. The nosological diagnosis was established in 96.5% of cases. Differential diagnosis was found to be difficult in 3.5% of cases.     

Certain nonlinear partially observable stochastic optimal controlproblems with explicit control laws equivalent to LEQG/LQG problems

This paper is concerned with partially observed stochastic optimal control problems when nonlinearities enter the dynamics of the unobservable state and the observations as gradients of potential functions. Explicit representations for the information state are derived in terms of a finite number of sufficient statistics. Consequently, the partially observed problem is recast as one of complete information with a new state generated by a modified version of the Kalman filter. When the terminal cost is quadratic in the unobservable state and includes the integral of the nonlinearities, the optimal control laws are explicitly computed, similar to linear-exponential-quadratic-Gaussian (LEQG) and linear-quadratic-Gaussian (LQG) tracking problems. The results are applicable to filtering and control of Hamiltonian systems     

Estimating the viewing parameters of random, noisy projections ofasymmetric objects for tomographic reconstruction

The ability to determine the viewing parameters of objects from their projections has enabled well established tomographic techniques to be employed in the 3D reconstruction of objects from images obtained via modalities where the orientation of the objects cannot be controlled. A method is described for the determination of the viewing parameters of randomly acquired projections of asymmetric objects. It extends upon the common lines algorithm by determining the relative orientation of projections from the location of lines of intersection among the Fourier transforms of the projections in 3D Fourier space. A new technique for finding the lines of intersection in the presence of translational displacement, and for subsequently finding the translational displacement, is presented. The complete algorithm is described and its efficacy is demonstrated using real data. A new technique for dealing with noise is also discussed     

On a large-gain theorem

The aim of this paper is to give a general quantitative requirement which the loop gain must satisfy in order to stabilize a given unstable (possibly nonlinear and time-varying) plant, namely that the gain must exceed one.     

Voluntary head stabilization in space during oscillatory trunk movements in the frontal plane performed in weightlessness

The ability voluntarily to stabilize the head in space during lateral rhythmic oscillations (0.59+/-0.09 Hz) of the trunk has been investigated during microgravity (microG) and normal gravity (nG) conditions (parabolic flights). Five healthy young subjects, who gave informed consent, were examined. The movements were performed with eyes open or eyes closed, during phases of either microG or nG. The main result was that head orientation with respect to vertical may be stabilized about the roll axis under microG with, as well as without vision, despite the reduction in vestibular afferent and muscle proprioceptive inputs. Moreover, the absence of head stabilization about the yaw axis confirms that the degrees of freedom of the neck can be independently controlled, as was previously reported. These results seem to indicate that voluntary head stabilization does not depend crucially upon static vestibular afferents. Head stabilization in space may in fact be organized on the basis of either dynamic vestibular afferents or a short-term memorized postural body schema.     

Histochemical evaluation of placental dipeptidyl peptidase IV (CD26) in pre-eclampsia: enzyme activity in villous trophoblast indicates an enhanced likelihood of gestational hypertensive disorders

Nitric oxide (NO) acts as a modulator of neuronal transmission in mature neuronal systems, including the retina. Recently, NO has also been suggested to have a trophic function during development. We examined immunocytochemically the distribution of NO-producing cells in developing and transplanted rabbit retinas. An antibody detecting the neuronal isoform of its biosynthetic enzyme, nitric oxide synthase (NOS), was used on normal developing retinas [starting at embryonic day (E) 15] and on rabbit retinal transplants after various survival times (1-139 days after surgery). Weakly stained cell bodies were first observed in the proximal margin of the neuroblastic layer at E 29. Stained processes projecting towards a developing inner plexiform layer were also visible at this time point. Immunoreactive cells were located at later stages in the innermost part of the inner nuclear layer and in the ganglion cell layer, and are likely to correspond mainly to amacrine cells. NOS-labelled cells were also found in retinal transplants. The first NOS-labelled cells appeared, as in normal developing retinas, in ages corresponding to E 29 and were still detected in transplants corresponding to postnatal day 123. NOS-labelled cells were seen in areas between rosettes, where amacrine cells are located. NOS-labelled processes were at times seen to project for long distances, forming very distinct plexuses. NOS-containing amacrine cells thus appear both in the transplants and in developing retinas in the embryonic stages, long before synaptic function involving these cells can be expected, suggesting a role for NO not only in neuromodulation but also in retinal development.     

Imprudent linking weaves a tangled Web

Hypertext linking is often embraced uncritically by Web authors eager to explore the power of hypertext without first considering its effects on their readers' comprehension. Hypertext linking is not a substitute for thought-we think with ideas, not with dissociated snippets of raw information. Even the most germane bits of information cannot become ideas, regardless of how cleverly they are stacked, listed or linked. Ideas define relevance, provide context and establish patterns. With patterns, most concepts become intelligible, and we need continuity and stability of theme and context to recognize patterns. So, like most powerful technologies, hypertext links are a mixed blessing. Used improperly, they can be detrimental to Web sites. “Loose links” can drive away an audience, dilute the site's message, confuse the reader with irrelevant digressions and become a continuing maintenance headache for site authors and Webmasters