Systems, devices, and methods related to a Deep Learning Accelerator and memory are described. For example, a digital camera may be configured to execute instructions with matrix operands and configured with: a housing; a lens; an image sensor positioned behind the lens to generate image data of a field of view of the digital camera; random access memory to store instructions executable by the Deep Learning Accelerator and store matrices of an Artificial Neural Network; a transceiver; and a controller configured to generate, and communicate using the transceiver to a separate computer, a description of an item or event in the field of view captured in the image data, based on an output of the Artificial Neural Network receiving the image data as an input. The separate computer may selectively request a portion of image data from the digital camera based on the processing of the description.

An image sensor includes a semiconductor substrate having first and second faces. The sensor includes a plurality of pixel groups each including pixels, each pixel having a photoelectric converter and a wiring pattern, the converter including a region whose major carriers are the same with charges to be accumulated in the photoelectric converter. The sensor also includes a microlenses which are located so that one microlens is arranged for each pixel group. The wiring patterns are located at a side of the first face, and the plurality of microlenses are located at a side of the second face. Light-incidence faces of the regions of the photoelectric converters of each pixel group are arranged along the second face such that the light-incidence faces are apart from each other in a direction along the second face.

The cost and power consumption of an imaging apparatus are reduced by facilitating detection of an incident angle of a light beam transmitted through a grating substrate. An image sensor converts an optical image captured by pixels arranged on an imaging surface and outputs the converted image signal. A modulator is configured to modulate intensity of light; and an image processing circuit performs image processing of the output image signal. The modulator has a grating substrate, a grating pattern formed on a back surface side of the grating substrate arranged in proximity to the light receiving surface of the image sensor; and a grating pattern formed on a front surface facing the back surface. Each of the grating patterns is constituted of a plurality of concentric circles. The modulator performs intensity modulation on the light transmitted through the grating pattern and outputs the modulated light to the image sensor.

A pixel PXL includes a first photodiode PDSL and a second photodiode PSLS having different well capacities and responsivities, transfer transistors TGSL-Tr, TGLS-Tr for transferring the charges stored in the photodiodes to a floating diffusion FD, and a capacitance changing part 80 for changing the capacitance of the floating diffusion depending on a capacitance changing signal. The first well capacity of the first photodiode PDSL is smaller than the second well capacity of the second photodiode PDLS, and the first responsivity of the first photodiode PDSL is larger than the second responsivity of the second photodiode PDLS. With these configurations, it becomes possible to realize a widened dynamic range, prevent the read-out noise from affecting the performance, and eventually achieve improved image quality.

In an image sensor, some pixels in an array contain a sampling circuit to sample the light intensity and a capacitor to store an analog value representing the intensity at that pixel. Alternatively, a group of pixel circuits will be equipped with such sampling and capacitor circuits. This allows simple redundancy-reducing computations with a relatively simple pixel architecture.

A display screen includes a plurality of pixels, wherein each pixel in the plurality of pixels includes at least one light-emitting element; and a plurality of light sensors, wherein each light sensor in the plurality of light sensors is disposed adjacent to a corresponding light-emitting element of a pixel in the plurality of pixels and is configured to convert received light into an electrical signal corresponding to a portion of an image.

A solid-state image sensor is provided. The solid-state image sensor includes a plurality of photoelectric conversion elements. The solid-state image sensor also includes a color filter layer disposed above the photoelectric conversion elements. The color filter layer has a plurality of color filter segments. The solid-state image sensor further includes a partition grid disposed between the color filter segments. Moreover, the solid-state image sensor includes a patterned structure disposed on the color filter layer. The patterned structure has a plurality of patterned segments. The solid-state image sensor also includes a transparent layer disposed on the color filter layer and the partition grid. The transparent layer surrounds the patterned segments. At least one patterned segment is disposed on the partition grid.

First and second readout circuits, each having a respective floating diffusion node, are coupled to a photodetection element within a pixel of an integrated-circuit image sensor. Following an exposure interval in which photocharge is accumulated within the photodetection element, a first portion of the accumulated photocharge is transferred from the photodetection element to the first floating diffusion node to enable generation of a first output signal within the first readout circuit, and a second portion of the accumulated photocharge is transferred from the photodetection element to the second floating diffusion node to enable generation of a second output signal within the second readout circuit. A digital pixel value is generated based on the first and second output signals.

A light sensing circuit and an image sensor are provided. The image sensor includes a first pixel unit including a plurality of first photodiodes and a first driving circuit to generate a first pixel signal based on an amount of charges stored in the plurality of first photodiodes, a second pixel unit including a plurality of second photodiodes and a second driving circuit to generate a second pixel signal based on an amount of charges stored in the plurality of second photodiodes, and a switching circuit connected to the first driving circuit through a first diffusion node and connected to the second driving circuit through a second diffusion node. The switching circuit connects or disconnects the first diffusion node and the second diffusion node based on a mode selection signal.

An imaging device with a plurality of image sensing pixels and a plurality of event detection pixels is provided. Each image sensing pixel includes a photoelectric conversion element and an imaging signal generation readout circuit. The image sensing readout circuit can be shared by a plurality of photoelectric conversion elements. Each event detection pixel includes a photoelectric conversion element and an event detection readout circuit. The event detection readout circuit can be shared by a plurality of photoelectric conversion elements. In addition, the photoelectric conversion element of an event detection pixel can be selectively connected to a shared imaging signal generation readout circuit. The number of image sensing pixels is greater than the number of event detection pixels. In addition, the area of a photoelectric conversion element of an event detection pixel can be greater than the area of a photoelectric conversion element of an image sensing pixel.