Search

Unraveling the quantitative properties of the spatio-temporal cortical responses to microstimulation is of essential importance for designing intracortical neural prostheses. Despite the millisecond-order precision of neuronal spiking and synaptic signaling in the cortex, the circuit response dynamics at such temporal resolution has been investigated to a limited extent. In the present, cortical neural responses to repetitive microstimulation pulses were examined in the mouse cerebral slices by means of the voltage-sensitive dye imaging. The amplitude of the population spikes induced directly by the stimulus pulses successively decreased when the inter-pulse interval of the stimuli was 5 msec, but not 20 msec. This was similar between the experiments in that the stimulation was applied to either the layer II/III or IV. These results suggested that the refractory period of spiking at the population level is one of limiting conditions for designing efficient frequency for repetitive pulse stimulation.

In visual prostheses, resolution of restored image remain a matter of concern. The most common strategy is mainly focused on enhancing the hardware capabilities of the prosthetic devices, e.g., the density of stimulating electrodes. However, it is not yet well understood how electrical stimulations delivered from multiple electrodes interact spatially and temporally. We conducted intracortical microstimulation in mice using a pair of glass electrodes inserted in the primary visual cortex and recorded the cortical responses using voltage-sensitive dye imaging, which allowed us a fairly good spatial and temporal resolution. Response induced by simultaneous stimulation of both electrodes was compared to linear off-line summation of individual stimulation responses. We found an enhancing effect of simultaneous stimulation in a weak intensity range (≤ 10 μA) and a suppressive effect in a high intensity range, compared to the linear summation of responses. Also, cortical responses to simultaneous stimulation were found to merge in the first tens of milliseconds after simultaneous stimulation on-set and gradually regress back to the stimulation sites. The results obtained in the present study provide a physiological basis for controlling electrical activities to improve spatial resolution of restored image, such as it is done in current steering.

Laser based displays have gathered much attention, and they are expected to make an alternative huge market for visible laser diodes (LDs), because only the displays may have a wide color gamut which can express ITU-R BT.2020. Single panel spatial light modulator (SLM) type projectors using broad area (BA) LDs for light sources are most popular so far, and multi-panel ones have started to spread especially for high luminosity. The laser light source for the projector consists of red, green and blue color LDs or blue LD + yellow phosphor (+ red LD). Common requests for the LDs are highly-reliable high-power operation at high case temperature.

Electrical microstimulation and near-infrared (NIR) laser light irradiation were combined to control neural activities in cerebral cortex in vitro. Spatio-temporal patterns of suprathreshold neural excitations in the primary visual cortex of mouse cerebral slices were visualized by the Ca2+-sensitive dye imaging. In response to a single pulse of the biphasic stimulus current delivered to the layer IV, neural excitations were initiated around the stimulating electrode tip, and then synaptically propagated to the layer II/III. The neural excitations and population excitatory post-synaptic potential in the layer II/III were inhibited in a spot region where NIR laser light was illuminated. This inhibitory effect was larger when the carbon-nanotube-bundle fiber as a light absorber was placed within the spot region. The present results suggested that the NIR laser light can be used to locally inhibit trans-synaptic neural excitations induced by electrical microstimulations.

This paper proposes a programmable stimulation controller for cortical visual prosthesis to configure spatio-temporal parameters of stimulation. Since the relationship between stimulation to the visual cortex and responses in vision has not been clarified enough both flexibility for stimulation strategies and timewise precision of stimulation are required for in-vivo experiments. In the proposed stimulation controller a 16bit microprocessor is utilized for various stimulation strategies and dedicated control signal generator for electrodes runs in parallel with the microprocessor. The proposed stimulation controller generates stimulations in temporal resolution of 1 yus and spatio resolution up to 4 096 stimulation sites. Evaluation with an FPGA demonstrates the programmability of its implementation.

The neuromorphic retina is a very large scale integrated (VLSI) circuit whose structure and function are designed to mimic those of retinal neuronal network. We have developed a neuromorphic retina model to study the function of retinal circuits in natural visual environment. The model consists of analog VLSI circuit and Field Programmable Gate Array (FPGA), enabling to reproduce responses of major types of retinal neuron in real time. Using the neuromorphic retina, we conducted “virtual in vivo experiments” in which neuronal images of retinal neurons responding to natural scenes are reproduced in real time. We also applied the neuromorphic retina to robotic vision and visual prosthetics. The neuromorphic device and system provide a novel methodology to advance research not only in electronic engineering but also in neuroscience.

Analyzing network architecture and spatio-temporal dynamics of the visual cortical areas can facilitate understanding visual information processing in the brain. Recently, several physiological experiments utilizing the fast in-vivo imaging technique have demonstrated that the primary visual cortex (V1) and the secondary visual cortex (V2) in mice exhibit complex properties of the responses to visual and electrical stimuli. In order to provide a tool for quantitatively analyzing such a complex dynamics of the cortices at the level of neurons and circuits, here, we constructed a physiologically plausible large-scale network model of the layers 2/3 of V1 and V2, composed of 14,056 multi-compartment neuron models. The Message-Passing-Interface-based parallel simulations of our network model were able to reproduce, at least quantitatively, the neural responses experimentally observed in mouse V1 and V2 with the voltage-sensitive dye imaging.

We developed a prototype very-large-scale integration chip of a multichannel current stimulator for stimulating neural tissues by utilizing 0.25 μm high-voltage complementary metal-oxide-semiconductor technology. Our designed chip has 20 output channels that are driven by five current buffers arranged in parallel; each buffer controls four output channels in time-sharing mode. The amplitude of a stimulation pulse can be controlled within a range of approximately ±100 μA/phase in each output channel. The stimulus parameters, e.g., amplitude and duration, are controlled separately for each channel by digital codes stored in built-in registers. Combinations of anode and cathode electrodes to pass the current can be changed online. We integrated our stimulator chip with a multielectrode array and studied the neuronal responses to multichannel current stimulations with various temporal patterns in mouse brain slices.

In this study, we developed a vision system that separates an image into a set of spatial frequency bands using multiple spatial filters during each single frame sampling period of 20 ms. The vision system comprises active pixel sensors (APS), two sample and hold (S/H) circuits, an analog resistive network, and a field-programmable gate array (FPGA). The resistive network facilitates instantaneous spatial filtering where its spatial property depends on its resistance. We implemented a digital circuit in the FPGA, which controls resistance of the resistive network and the timing when reading out spatial-filtered images in a single frame. We examined the performance of this system by presenting a set of test images. The system output three spatial band-pass filtered images at 50 fps.

We assessed the feasibility of applying a resistive network (RN) filter to the scale-invariant feature transform (SIFT) algorithm by performing computer simulations for the hardware implementation of the filter. SIFT is an algorithm for computer vision to describe and detect local features that are invariant to scale and rotation of objects. However, it is difficult to perform multiple spatial filterings in SIFT algorithm in real time due to its high computational cost. To solve this problem, we employed an RN which performs spatial filtering instantaneously with extremely low power dissipation. In order to apply an RN filter to the SIFT algorithm instead of Gaussian filter, which is employed in the original SIFT algorithm, we investigated the difference in the spatial properties of the two filters. We simulated the SIFT algorithm employing the RN filter on a computer, and we demonstrated that key points were detected at the same place irrespective of the image size, and that the scale of the key point was detected appropriately.

We describe the reliable vergence eye movement control of a binocular robot vision system based on a disparity computation in the primary visual cortex (V1). The system consists of two silicon retinas, simple cell chips, and an FPGA. The silicon retinas emulate a Laplacian-Gaussian (▽2G)-like receptive field of the vertebrate retina. The simple cell chips generate an orientation-selective receptive field by aggregating multiple pixels of the silicon retina, mimicking the Hubel and Wiesel type feed-forward model. The FPGA receives the outputs from the two simple cell chips and calculates the responses of complex cells based on the disparity energy model. This system provides complex cell outputs tuned to five different disparities in real-time. The vergence control signal is obtained by pooling these multiple complex cell responses. We thus demonstrate that the present system can robustly execute vergence movement even in a complex scene.

High power 625-nm AlGaInP laser diode was fabricated and evaluated. Remarkable short wavelength lasing at 624.9 nm was achieved on the condition of T c = 25°C and 50 mW output under CW operation. At injection current of 800 mA, a high output power of 220 mW was obtained. Conversion efficiency from electric power input to luminous flux for 10 lm output was calculated as 11 lm/W. In order to apply a 625-nm LD to practical display, further improvement of its temperature characteristics is necessary.

In physiological studies conducted thus far, the functional roles of visual neurons have been discussed using models based on the responses of a limited number of neurons to simple visual stimuli. However, these discussions based on simplified inputs do not always hold in the case of highly complex visual environments wherein actual biological vision systems operate. In order to examine and discuss the visual functions of the models, a system that can emulate the responses of a number of model neurons in real time is necessary. Therefore, we have developed a system for the real-time reconstruction of neural images with physiologically reasonable spatiotemporal properties using an analog-digital hybrid system consisting of analog resistive networks, field-programmable gate array (FPGA) circuits, and a digital computer. Here, we conducted emulations for natural scenes with 128 × 128 sustained/transient retinal neurons at 60 frames per second.

The high-power operation of the lateral single mode 638 nm AlGaInP laser diode was demonstrated. The stable operation up to 220 mW at 55degC was realized. This is the highest power record among the narrow stripe 638-nm LDs. This LD is suitable for micro-projector.

A silicon retina is an intelligent vision sensor that can execute real-time image pre-processing by using a parallel analog circuit that mimics the structure of the neuronal circuits in a vertebrate retina. In order to enhance robustness against changes in lighting conditions, we designed and fabricated a frame-based, wide dynamic range silicon retina with an active pixel sensor that approximates the logarithmic illumination-to-voltage transfer characteristics by using a time-dependent stepped reset voltage technique. The chip in this study realized dynamic range wide enough for perceiving objects in both indoor and outdoor environments. Moreover, the combination of the logarithmic-like photosensor and the Laplacian-Gaussian-like filtering by a resistive network in the chip produces the response with lightness constancy, that is, the response of the chip depends only on the contrast of the reflectance of objects, and not on the changes in illumination. The present silicon retina is suitable for the front end of the vision system in autonomous mobile robots in the real world.

We describe a multi-chip analog VLSI vision sensor consisting of a silicon retina and an orientation chip that is tuned to particular orientations and spatial frequency bands of input image. The silicon retina emulates the neural circuits of the vertebrate retina and exhibits a Laplacian-Gaussian-like receptive field. The orientation chip receives output image of the silicon retina and aggregates multiple pixels in the image to generate a Gabor-like orientation-selective spatial filter. The orientation and spatial frequency properties of the filter is controlled by selecting the pixels to be aggregated using external signals. Multiple orientations and spatial frequency bands were implemented in separate chips that are connected in parallel to the silicon retina to obtain output images tuned to different orientations and spatial frequency bands simultaneously.

A high resolution neuromorphic multi-chip vision system was fabricated to emulate the orientation selective response of the simple cell in the primary visual cortex. The vision system consists of two types of analog chips: a silicon retina and a simple cell chip. The center-surround concentric receptive fields of the silicon retina are aggregated in the simple cell chip, mimicking the hierarchical architecture in the visual system of the brain. Both chips have 100 /spl times/ 100 pixels. Using the orientation-selective outputs obtained from the multi-chip system, texture segregation was conducted based on a computational model inspired by neurophysiology. The texture image was filtered by the two orthogonally oriented receptive fields of the multi-chip system and the filtered images were combined to segregate the area of different texture orientation with the aid of field programmable gate array (FPGA). The present multi-chip system is useful to emulate and verify computational models for texture segregation of the cortical cells.

In this research, a dialog environment between human and computers has been constructed. This dialog environment is based on the integration between verbal information through the utterances and non-verbal information by the gestures such as finger pointing. The experimental results have proved the effectiveness of this approach in terms of facilitating man-machine interaction and communication. Currently, keyboard and mouse are widely used as the interface to the electronic devices such as computers. Elderly people cannot use them very easily. In Japan, if various apparatus is connected along with the arrival of ubiquitous network society, people have to handle complicated operations. The reality is that even people exploiting such operations feel it difficult to use new ones while growing older. The environment constructed in this research allows a user to communicate by talking and showing gestures to a personified agent who is a salesman selling fountain pens in a virtual shop. A user can use his/her finger to point at a pen and ask a question such as 'Would you show me this pen?' Through the integration of gesture interpretation and verbal information analysis, it is possible for user to communicate the pen he wants to the virtual agent. If the fingertip is too far way from the pen, it may be difficult for the agent to tell which pen the user wants. This needs the further communication through utterance or gestures or both.

We describe a multi-chip analog VLSI system which emulates the orientation selective response of the simple cell in the primary visual cortex. The system consists of two analog chips, a silicon retina and an orientation chip, which mimic the parallel and hierarchical architecture of the visual system in the brain. The spatial resolution of the system is 100/spl times/100 pixels. The silicon retina receives the input image with a Laplacian-Gaussian-like receptive field. The output image of the retina is transferred to the orientation selection chip. The image transmission between two chips is carried out using analog signals. The orientation chip selectively aggregates multiple pixels of the silicon retina, mimicking the feed-forward model proposed to explain the orientation-selective receptive field of the simple cell. The orientation chip provides three output images with different preferred orientation in parallel. The analog multi-chip architecture used in the present study is useful to mimic the responses of higher order cells in the primary visual cortex.

A 100/spl times/100 pixel analog silicon retina was fabricated. The two functions of the silicon retina were verified: a Laplacian-Gaussian-like spatial filtering and a subtraction of consecutive frames. The analog silicon retina fabricated in the present study has matured from a development prototype to full-scale practical usefulness. Furthermore, we have designed a pre-processing system that consists of the silicon retina and FPGA. The system has compact hardware and low power dissipation and therefore is suitable for robotic vision.

With the development of data processing technology, the number of people using a computer is rapidly growing now. And they came to use it in the everyday life of ordinary homes. However, it is hard for elderly people and disabled persons to treat an interface device such as a mouse and a keyboard. For this reason, the dialog system is needed that permits them to dialog with a computer. A virtual human called avatar is introduced into virtual space. Users can talk to him as if he talked to a real person. When a man talks in real space, the partner chimes in according to a speaker's utterance. By synchronizing utterance with a behavior, a partner can understand a speaker's intention correctly. Moreover, relation between a listener and a speaker is not always fixed in man's conversation, and the relation interchanges. This research proposes how to understand the conversation accompanied with the pointing behavior

We have proposed a new method of suppressing the lasing-wavelength-change by means of decreasing facet reflectivity as increasing wavelength. Less than one-third reduction of losing wavelength was realized by this method.

Numerical analysis on the influence of the structural asymmetry on the vertical beam quality and optimization of laser characteristics are studied in the paper. Deflection of the vertical beam toward the GaAs substrate is caused by the disagreement of the steady-state optical field between the active region and the window region. The calculated angle of deflection is approximately 1.5 /spl deg/, which agrees with the measurement.

A silicon retina was fabricated to emulate two fundamental types of response in the vertebrate retinal circuit, namely the sustained and transient responses. The sustained response exhibits a Laplacian-Gaussian-like receptive field. The transient response is obtained by subtracting consecutive image frames. The outputs of the chip are offset-suppressed analog voltages since uncontrollable mismatches of transistor characteristics are compensated for with the aid of sample/hold circuits embedded in each pixel circuit. The chip was applied to extract direction of motion using FPGA in real-time under an indoor illumination.

A kink mechanism in 660-nm laser diodes (LDs) has been studied experimentally. The experiments revealed that the main origin of the kink is a refractive index change due to heat generation in the stripe portion, and the kink power can be increased by improving the temperature characteristics of the LD. A newly developed LD, based on this result, shows stable lateral mode operation up to 190 mW at 80/spl deg/C. This is the highest power recorded among narrow stripe LDs with a wavelength of 660 nm. This LD is suitable for the next generation of high-speed (8x-) DVD-R/RW drives necessitating 140 mW class LDs.

A linear analog network model is proposed to describe the neural circuit of the outer retina. Using the model, we characterized the spatial filtering properties of the circuit in terms of the standard regularization theory in which the early vision problems are attributed to minimization of a cost function. Inspired by the algorithm/architecture of the circuit, a vision chip was designed using analog CMOS VLSI circuit technology. The chip fabricated is an one-dimensional 100 pixel line sensor. In the chip, sample/hold amplifier circuits are incorporated to compensate for the fixed pattern noise of photosensors as well as statistic transistor mismatch. The chip effectively extracts contour of given images by controlling the receptive field size, simulating a light-adaptive mechanism suggested in the vertebrate retinal circuit.

The high power operation of the lateral mode stabilized 785nm AlGaAs LD with the window-mirror structure has been demonstrated. The stable lateral mode operation up to 250mW (kink level of 280 mW) is realized. To the best our knowledge, this is the highest power record among the narrow stripe LDs with a wavelength of 785 nm and is suitable for CD ROM disc drives.