We designed a CMOS image sensor capable of capturing variably smoothed images. This sensor uses a negative-feedback technique to set photodiode (PD) capacitance in the pixel circuit to any intermediate voltage during charge accumulation and it provides a neighboring-pixel operation by using their average value when resetting the PD capacitance. Smoothing-filter coefficients are changed by adjusting timing of the pixel-readout and neighboring-pixels operations. The performance of the proposed sensor was evaluated by SPICE simulation and numerical analysis.

Deep neural network (NN) has been widely accepted for enabling various AI applications, however, the limitation of computational and memory resources is a major problem on mobile devices. Quantized NN with a reduced bit precision is an effective solution, which relaxes the resource requirements, but the accuracy degradation due to its numerical approximation is another problem. We propose a novel quantized NN model employing the “dithering” technique to improve the accuracy with the minimal additional hardware requirement at the view point of the hardware-algorithm co-designing. Dithering distributes the quantization error occurring at each pixel (neuron) spatially so that the total information loss of the plane would be minimized. The experiment we conducted using the software-based accuracy evaluation and FPGA-based hardware resource estimation proved the effectiveness and efficiency of the concept of an NN model with dithering.

We have developed a Si-CMOS terahertz image sensor to address the paucity of low-cost terahertz detectors. Our imaging pixel directly connects to a VCO-based ADC and achieves pixel parallel ADC architecture for high-speed global shutter THz imaging. In this paper, we propose a digital calibration technique for offset and gain variation of each pixel using global shutter operation. The calibration technique gives reference signal to all pixels simultaneously and takes reference frames as a part of the high-speed image captures. Using this technique, we achieve offset/non-linear gain variation suppression of 85.7% compared to without correction.

An annealing processor based on the Ising model is a remarkable candidate for combinatorial optimization problems and it is superior to general von Neumann computers. CMOS-based implementations of the annealing processor are efficient and feasible based on current semiconductor technology. However, critical problems with annealing processors remain. There are few simulated spins and inflexibility in terms of implementable graph topology due to hardware constraints. A prior approach to overcoming these problems is to emulate a complicated graph on a simple and high-density spin array with so-called minor embedding, a spin duplication method based on graph theory. When a complicated graph is embedded on such hardware, numerous spins are consumed to represent high-degree spins by combining multiple low-degree spins. In addition to the number of spins, the quality of solutions decreases as a result of dummy strong connections between the duplicated spins. Thus, the approach cannot handle large-scale practical problems. This paper proposes a flexible and scalable hardware architecture with time-division multiplexing for massive spins and high-degree topologies. A target graph is separated and mapped onto multiple virtual planes, and each plane is subject to interleaved simulation with time-division processing. Therefore, the behavior of high-degree spins is efficiently emulated over time, so that no dummy strong connections are required, and the solution quality is accordingly improved. We implemented a prototype hardware design for FPGAs, and we evaluated the proposed method in a software-based annealing processor simulator. The results indicate that the method increased the spins that can be deployed. In addition, our time-division multiplexing architecture improved the solution quality and convergence time with reasonable resource consumption.

For a digitally controlled phase-locked loop (PLL), we evaluate the use of a clock-period comparator (CPC). In this PLL, only the frequency lock operation should be performed; however, the phase lock operation is also simultaneously achieved by performing the clock-period comparison when the phases of the reference signal and the output signal approach each other. Theoretically a lock-up operation was conducted. In addition, we succeeded in digitizing a voltage controlled oscillator (VCO) with a linear characteristic. We confirmed a phase lock operation with a slight loop characteristic through SPICE simulation.