Abstract: A prosthetic retina for implantation in an eye having a defective retina is formed from an array of nanowires having a predetermined spatial distribution, density, size and shape implanted in close proximity to the retina. An electrical conductor is formed at a first end of all nanowires in the array of nanowires and placed in contact with a bias source which biases the array. A plurality of electrodes is located on a second end of each of one nanowire or a bundle of nanowires in the array. Each nanowire produces a photocurrent at a corresponding electrode in response to detection of light impinging on the array of nanowires and the photocurrent stimulates one or more neurons adapted for visual perception. In the preferred embodiment, the predetermined spatial distribution mimics a distribution of rods and cones in a normal eye.

Abstract: An electric field sensor includes one or more sensing electrodes connected to an integrated amplifier that bootstraps all parasitic capacitances at the sensor input to provide for a very high input impedance without the need for neutralization or other adjustments and calibration. The integrated amplifier for the electric field sensor further includes low-noise ESD/biasing structures to stabilize the DC-potential of the sensor with a minimum amount of added noise, leakage and parasitic capacitance.

Abstract: An electric field sensor includes one or more sensing electrodes connected to an integrated amplifier that bootstraps all parasitic capacitances at the sensor input to provide for a very high input impedance without the need for neutralization or other adjustments and calibration. The integrated amplifier for the electric field sensor further includes low-noise ESD/biasing structures to stabilize the DC-potential of the sensor with a minimum amount of added noise, leakage and parasitic capacitance.

Abstract: A real-time multi-channel EEG signal processor based on an on-line recursive independent component analysis is provided. A whitening unit generates covariance matrix by computing covariance according to a received sampling signal. A covariance matrix generates a whitening matrix by a computation of an inverse square root matrix calculation unit. An ORICA calculation unit computes the sampling signal and the whitening matrix to obtain a post-whitening sampling signal. The post-whitening sampling signal and an unmixing matrix implement an independent component analysis computation to obtain an independent component data. An ORICA training unit implements training of the unmixing matrix according to the independent component data to generate a new unmixing matrix. The ORICA calculation unit may use the new unmixing matrix to implement an independent component analysis computation. Hardware complexity and power consumption can be reduced by sharing registers and arithmetic calculation units.

Abstract: A prosthetic retina for implantation in an eye having a defective retina is formed from an array of nanowires having a predetermined spatial distribution, density, size and shape implanted in close proximity to the retina. An electrical conductor is formed at a first end of all nanowires in the array of nanowires and placed in contact with a bias source which biases the array. A plurality of electrodes is located on a second end of each of one nanowire or a bundle of nanowires in the array. Each nanowire produces a photocurrent at a corresponding electrode in response to detection of light impinging on the array of nanowires and the photocurrent stimulates one or more neurons adapted for visual perception. In the preferred embodiment, the predetermined spatial distribution mimics a distribution of rods and cones in a normal eye.

Abstract: A capacitive sensor system including a sensing plate, an amplifier, and a switching circuit is described. The sensing plate is capacitively coupled to a body surface. A change in the electric potential on the body surface generates an electric field that induces change in the electric potential of the sensing plate. The sensing plate includes a sensing node positioned in the electric field for generating an input signal from the electric field. The sensing plate is not in contact with the body surface. The amplifier receives the input signal at the input port, amplifies the input signal and generates an output signal at the output port. The switching circuit is connected to the input port and a reference voltage. The switching circuit non-continuously closes a shunting path from the sensing node to the reference voltage to reset the voltage at the sensing node.

Abstract: An electric field sensor includes one or more sensing electrodes connected to an integrated amplifier that bootstraps all parasitic capacitances at the sensor input to provide for a very high input impedance without the need for neutralization or other adjustments and calibration. The integrated amplifier for the electric field sensor further includes low-noise ESD/biasing structures to stabilize the DC-potential of the sensor with a minimum amount of added noise, leakage and parasitic capacitance.

Abstract: A network of coupled neurons for implementing Non-Lipschitz dynamics for modeling nonlinear processes or conditions comprising: a plurality of neurons, each being configurable in attractor and repulsion modes of operation, and programmable by an external signal; a plurality of synaptic connections for connecting at least a portion of the plurality of neurons for passage of data from one neuron to another; feedback circuitry for incrementing and decrementing an analog voltage output depending upon the output of the synaptic connection; whereby by the circuit solves Non-Lipschitz problems by programmably controlling the attractor and repulsion modes.

Abstract: A capacitive sensor system including a sensing plate, an amplifier, and a switching circuit is described. The sensing plate is capacitively coupled to a body surface. A change in the electric potential on the body surface generates an electric field that induces change in the electric potential of the sensing plate. The sensing plate includes a sensing node positioned in the electric field for generating an input signal from the electric field. The sensing plate is not in contact with the body surface. The amplifier receives the input signal at the input port, amplifies the input signal and generates an output signal at the output port. The switching circuit is connected to the input port and a reference voltage. The switching circuit non-continuously closes a shunting path from the sensing node to the reference voltage to reset the voltage at the sensing node.

Abstract: Analog computational arrays for matrix-vector multiplication offer very large integration density and throughput as, for instance, needed for real-time signal processing in video. Despite the success of adaptive algorithms and architectures in reducing the effect of analog component mismatch and noise on system performance, the precision and repeatability of analog VLSI computation under process and environmental variations is inadequate for some applications. Digital implementation offers absolute precision limited only by wordlength, but at the cost of significantly larger silicon area and power dissipation compared with dedicated, fine-grain parallel analog implementation. The present invention comprises a hybrid analog and digital technology for fast and accurate computing of a product of a long vector (thousands of dimensions) with a large matrix (thousands of rows and columns).

Abstract: A method of separating and localizing sources of traveling waves, by obtaining linearly independent time-differentiated instantaneous observations of the sources, from spatial derivatives of the traveling wave acquired using a distributed sensor or a sensor array. The sources are blindly separated by direct application of (static) independent component analysis on the time-differentiated observations, yielding both the sources and their direction cosines relative to the sensor geometry. The method is suited for arrays of small aperture, with dimensions shorter than the coherence length of the waves. In one preferred embodiment, three sources are separated and localized from differential observations on four coplanar sensors positioned on the corners of a square.

Abstract: A method of separating and localizing sources of traveling waves, by obtaining linearly independent time-differentiated instantaneous observations of the sources, from spatial derivatives of the traveling wave acquired using a distributed sensor or a sensor array. The sources are blindly separated by direct application of (static) independent component analysis on the time-differentiated observations, yielding both the sources and their direction cosines relative to the sensor geometry. The method is suited for arrays of small aperture, with dimensions shorter than the coherence length of the waves. In one preferred embodiment, three sources are separated and localized from differential observations on four coplanar sensors positioned on the corners of a square.

Abstract: Method and apparatus for using current-mode analog circuits to compute the correlation between an auditory input signal in the time-frequency domain and a stored binary template. By using massively parallel computation, circuits made of relatively imprecise analog components are able to accurately classify transient events, are competitive with high-performance DSP systems, and operate with much smaller power requirements.

Abstract: A method and apparatus is described for refreshing the analog content of a analog memory having a volatile storage device. An analog value stored on a volatile storage device is iteratively adjusted to maintain the analog value in proximity to one of a set of predetermined discrete analog memory levels. Binary quantization of the stored value, yielding one bit of information corresponding to the analog value stored, determines whether to increase or decrease the stored value by a given small amount. In essence, the bit obtained by binary quantization encodes the direction toward the nearest discrete level. Memory retention achieved by periodic iteration of the method is robust to noise and random errors in the quantization, over a wide range of operating conditions.

Abstract: The present invention provides a compact and robust architecture and a corresponding method to implement a monotonic algorithmic D/A converter that processes the bits of the digital input in the order from MSB to LSB, and a successive approximation A/D converter employing the intermediate conversion results of this D/A converter. The invention is aimed at applications requiring a dense integration in general VLSI technologies of multiple D/A and A/D converters, where individual trimming of components to compensate for component offsets and mismatches is virtually impossible. The architecture comprises four charge holding components, one switch for charge sharing, two bi-directional replication elements for charge storage and recall, and one comparator. Also described is an efficient way of performing pseudo-logarithmic compression of conversion values merely by adjusting the relative sizes of two of the charge holding components.