Apparatuses, systems, and techniques are presented to train one or more neural networks. In at least one embodiment, one or more neural networks are trained based, at least in part, on one or more image sequences, where backpropagation is performed using one or more subsets of images from the one or more image sequences.

A method for generating a super resolution image may comprise up-scaling an input low resolution image; determining a directivity for each patch included in the up-scaled image; selecting an orientation-specified neural network or an orientation-non-specified neural network according to the directivity of the patch; applying the selected neural network to the patch; and obtaining a super resolution image by combining one or more patches output from the orientation-specified neural network and the orientation-non-specified neural network.

Methods, devices and non-transitory computer-readable storage medium for processing a sequence of respective top view images of a same terrestrial location are provided. One of the method may comprise choosing one image, called reference image, among the respective top view images, estimating for each respective top view image a respective geometric deformation between the respective top view image and the reference image, computing by the respective geometric deformations respective subpixel positions of the respective top view images relative to one high-resolution coordinate system, interpolating at the respective subpixel positions to sample at least part of at least some of the respective top view images on a prescribed grid to obtain a high-resolution image.

Presented herein are embodiments of systems and methods for training a system and using a trained system to generate super-resolution imagery from low-resolution imagery. Embodiments for generating super-resolution imagery from low-resolution imagery include obtaining an input trade-off parameter that indicates a preference regarding low distortion or high perceptual quality for a generated SR image and obtaining a latent variable from a distribution defined, at least in part, by a trade-off parameter. Embodiments include inputting an LR image and the latent variable into an embodiment of an invertible rescaling network (IRNN) in an inverse upscaling direction of the IRNN to generate an output SR image that comprises accuracy and perception qualities conditioned by the input trade-off parameter. In one or more embodiments, a trained IRNN accepts a value for a trade-off parameter that indicates a desired trade-off between whether the trained IRNN generates an SR image having lower distortion or higher perceptual quality.

An electronic device includes at least one imaging sensor and at least one processor coupled to the at least one imaging sensor. The at least one imaging sensor is configured to capture a burst of image frames. The at least one processor is configured to generate a low-resolution image from the burst of image frames. The at least one processor is also configured to estimate a blur kernel based on the burst of image frames. The at least one processor is further configured to perform deconvolution on the low-resolution image using the blur kernel to generate a deconvolved image. In addition, the at least one processor is configured to generate a high-resolution image using super resolution (SR) on the deconvolved image.

The present disclosure provides a picture processing method and device, including: an integrated circuit chip IC receiving a to-be-processed picture sent by a graphics processor GPU; the IC pre-processing the to-be-processed picture; the IC performing counter-distortion process on the pre-processed picture; and the IC outputting the picture which is subjected to the counter-distortion process for display.

A system for generating a high resolution (HR) computed tomography (CT) image from a low resolution (LR) CT image is described. The system includes a first generative adversarial network (GAN) and a second GAN. The first GAN includes a first generative neural network (G) configured to receive a training LR image dataset and to generate a corresponding estimated HR image dataset, and a first discriminative neural network (D.sub.Y) configured to compare a training HR image dataset and the estimated HR image dataset. The second GAN includes a second generative neural network (F) configured to receive the training HR image dataset and to generate a corresponding estimated LR image dataset, and a second discriminative neural network (D.sub.X) configured to compare the training LR image dataset and the estimated LR image dataset. The system further includes an optimization module configured to determine an optimization function based, at least in part, on at least one of the estimated HR image dataset and/or the estimated LR image dataset. The optimization function contains at least one loss function. The optimization module is further configured to adjust a plurality of neural network parameters associated with at least one of the first GAN and/or the second GAN, to optimize the optimization function.

An imaging system for a vehicle for obtaining an anti-flickering super-resolution image includes an image sensor adapted to obtain a sequence of images, and an image processor adapted to receive the sequence of images, compare image information of a most recent image of the sequence of images to a reference image to detect at least one image region of mismatch in the most recent image, remove the detected image region from image information of the most recent image to obtain adjusted image information, and add the adjusted image information of the most recent image to a super-resolution image.

A microscopy method includes a trained deep neural network that is executed by software using one or more processors of a computing device, the trained deep neural network trained with a training set of images comprising co-registered pairs of high-resolution microscopy images or image patches of a sample and their corresponding low-resolution microscopy images or image patches of the same sample. A microscopy input image of a sample to be imaged is input to the trained deep neural network which rapidly outputs an output image of the sample, the output image having improved one or more of spatial resolution, depth-of-field, signal-to-noise ratio, and/or image contrast.

A method for generating a super-resolution image and related device is provided. In one aspect, the method comprises: receiving a first low-resolution image and a second low-resolution image, the first low-resolution image and second low-resolution image have a first spatial resolution and having been captured simultaneously by a pair of pixel arrays of a common image sensor, wherein the pixel arrays of the image sensor are located as to be diagonally shifted from each other by a sub-pixel increment; adaptively enhancing the first low-resolution and the second low-resolution images to generate an enhanced first low-resolution image and an enhanced second low-resolution image, respectively; mapping (e.g., non-uniformly) pixels of each of the enhanced first and second low-resolution images to a super-resolution grid having a spatial resolution greater than the first spatial resolution to generate a first intermediate super-resolution image and a second intermediate super-resolution image, respectively; and combining the first intermediate super-resolution image and second intermediate super-resolution image to generate a composite super-resolution image.