A neural network-based rendering technique increases temporal stability and image fidelity of low sample count path tracing by optimizing a distribution of samples for rendering each image in a sequence. A sample predictor neural network learns spatio-temporal sampling strategies such as placing more samples in dis-occluded regions and tracking specular highlights. Temporal feedback enables a denoiser neural network to boost the effective input sample count and increases temporal stability. The initial uniform sampling step typically present in adaptive sampling algorithms is not needed. The sample predictor and denoiser operate at interactive rates to achieve significantly improved image quality and temporal stability compared with conventional adaptive sampling techniques.

A method of enhancing an x-ray image is disclosed. The method involves obtaining an input image based on a source x-ray image of an object. Compositional information representing physical characteristics of the object is also obtained. An image enhancement process is applied to the input image to generate a processed image. Application of the image enhancement process is controlled by one or more parameters determined in dependence on the compositional information. An output image is then provided based on the processed image.

A character segmentation method and apparatus, and a computer-readable storage medium. The character segmentation method comprises: converting a character area image into a grayscale image (step 101); converting the grayscale image into an edge binary image by using an edge detection algorithm (step 102); acquiring character box segmentation blocks from the edge binary image by using a projection method (step 103); and determining a target character area from the character box segmentation blocks by using a contour detection algorithm, and performing character segmentation on the character area image according to the target character area (step 104). Another character segmentation method comprises: converting a character area image into a grayscale image (step 701); performing clustering analysis on the grayscale image by using a fuzzy C-means clustering algorithm, and executing binarization processing on the grayscale image according to a clustering analysis result (step 702); acquiring character positioning blocks from a binary image by using a projection method (step 703); and performing character segmentation on the character area image according to position information of the character positioning blocks (step 704). By using the methods and apparatuses, character segmentation can be performed on a relatively low quality image.

The present disclosure provides a method for incrementing a sample image, an electronic device, and a computer readable storage medium. A specific implementation comprises: acquiring a first convolutional feature of an original sample image; determining, according to a region generation network and the first convolutional feature, a candidate region and a first probability that the candidate region contains a target object; determining a target candidate region from the candidate region based on the first probability, and mapping the target candidate region back to the original sample image to obtain an intermediate image; and performing image enhancement processing on a portion of the intermediate image corresponding to the target candidate region and/or performing image blur processing on a portion of the intermediate image corresponding to a non-target candidate region to obtain an incremental sample image.

The invention provides for a medical apparatus (100, 300, 400) comprising a subject support (102) configured for moving a subject (106) from a first position (124) to a second position (130) along a linear path (134). The subject support comprises a support surface (108) for receiving the subject. The subject support is further configured for positioning the subject support in at least one intermediate position (128). The subject support is configured for measuring a displacement (132) along the linear path between the first position and the at least one intermediate position. Each of the at least one intermediate position is located between the first position and the second position. The medical apparatus further comprises a camera (110) configured for imaging the support surface in the first position. Execution of machine executable instructions 116 cause the a processor (116) controlling the medical apparatus to: acquire (200) an initial image (142) with the camera when the subject support is in the first position; control (202) the subject support to move the subject support from the first position to the second position; acquire (204) at least one intermediate image (144) with the camera and the displacement for each of the at least one intermediate image as the subject support is moved from the first position to the second position; and calculate (206) a height profile (150, 600, 604) of the subject by comparing the initial image and the at least one intermediate image. The height profile is at least partially calculated using the displacement. The height profile is descriptive of the spatially dependent height of the subject above the support surface.

A recording device provides bystander-centric privacy controls for authorizing the storage of a bystander's identifying information (e.g., video or audio recordings of the bystander). Before a recording device can store identifying information of bystanders, the bystanders may indicate to the recording device whether they authorize the storage. If the bystanders do not authorize the storage, the recording device may modify the identifying information captured by sensors, such as a video camera or a microphone, such that the identity of the non-authorizing bystander is not identifiable through the modified identifying information. Thus, bystanders are given increased agency over whether they want to be recorded. Further, if the bystanders do not want to be recorded, sensor data that may identify them is modified by the recording device to prevent unwanted exposure of their identity in recorded content.

A facial verification method includes separating a query face image into color channel images of different color channels, obtaining a multi-color channel target face image with a reduced shading of the query face image based on a smoothed image and a gradient image of each of the color channel images, extracting a face feature from the multi-color channel target face image, and determining whether face verification is successful based on the extracted face feature.

Provided are methods, performed by an electronic device, for processing an image. The method includes obtaining a first image by photographing a subject. The method further includes obtaining a depth image including information related to a distance from the electronic device to the subject. The method further includes determining whether light reflection exists in the first image. The method further includes obtaining depth information indicating the distance from the electronic device to the subject. The method further includes obtaining a second image by photographing the subject in an activated state of a flash. The method further includes performing pre-processing for matching the first image, the second image, and the depth image. The method further includes obtaining the image from which the light reflection has been removed using at least one of the pre-processed first image, the pre-processed second image, or the pre-processed depth image.

An electronic device may include enhancement circuitry to enhance high resolution image data to improve perceived quality of an image corresponding to the high resolution image data. The enhancement circuitry may include tone detection circuitry to determine one or more tones within the image and apply changes to the high resolution image data based on the one or more tones. The enhancement circuitry may also include example-based improvement circuitry to compare the high resolution image data to low resolution image data and apply changes to the high resolution image data based on differences between sections of the high resolution image data and sections of the low resolution image data. The enhancement circuitry may also include channel processing circuitry to apply the first and second changes to one or more channels of the high resolution image data.

One embodiment provides a method comprising receiving an input content, and receiving ambient contextual data indicative of one or more ambient lighting conditions of an environment including a display device. The input content has corresponding metadata that at least partially represents a creative intent indicative of how the input content is intended to be viewed. The method further comprises adaptively correcting the input content based on the ambient contextual data to preserve the creative intent, and providing the corrected input content to the display device for presentation. The adaptively correcting comprises applying automatic white balancing to the input content to correct color tone of the input content.