Abstract: Apparatuses, systems, and techniques to use one or more neural networks to generate an upsampled version of one or more images based, at least in part, on a denoised version of said one or more images. At least one embodiment pertains to generating an upsampled high-resolution image from a noisy version and denoised version of a low-resolution image. At least one embodiment pertains to separating components of a low-resolution image before denoising an image.

Abstract: The present application provides an input device. The input device comprises a key module, a capacitance detection module and a control module. The key module receives a physical pressing signal and correspondingly feeds back a capacitance change signal, the capacitance detection module receives the capacitance change signal from the key module and determines the working state of the key module, and the control module receives the determination result from the capacitance detection module and correspondingly generates an input signal. The key module comprises a key switch and a capacitor circuit arranged in series. The key module responds to the external physical pressing signal, and correspondingly adjusts the capacitance value change of the key module. By serially connecting a capacitor circuit, the sensitivity of determining the working state of a key module is improved, the requirements for assembly precision of a key switch are reduced, and the installation difficulty of products is reduced.

Abstract: Disclosed are an electronic cigarette and a vaporizer thereof which are suitable for adjusting air inflow with a single hand. The electronic cigarette vaporizer includes a vaporizer base and a liquid reservoir connected to the vaporizer. An atomizing cavity is formed by an internal cavity obtained by the connection of the liquid reservoir and the vaporizer base. An air inlet passage communicated with the atomizing cavity is arranged at the liquid reservoir. An adjusting knob at the entrance of the air inlet passage for adjusting air inflow is also arranged at the liquid reservoir. An air inlet is arranged at the adjusting knob. By rotating the adjusting knob, the air inflow can be adjusted by relative position changes of the air inlet and the entrance of the air inlet passage.

Abstract: The disclosure provides a learning framework that unifies both semantic segmentation and semantic edge detection. A learnable recurrent message passing layer is disclosed where semantic edges are considered as explicitly learned gating signals to refine segmentation and improve dense prediction quality by finding compact structures for message paths. The disclosure includes a method for coupled segmentation and edge learning. In one example, the method includes: (1) receiving an input image, (2) generating, from the input image, a semantic feature map, an affinity map, and a semantic edge map from a single backbone network of a convolutional neural network (CNN), and (3) producing a refined semantic feature map by smoothing pixels of the semantic feature map using spatial propagation, and controlling the smoothing using both affinity values from the affinity map and edge values from the semantic edge map.

Abstract: Apparatuses, systems, and techniques for texture synthesis from small input textures in images using convolutional neural networks. In at least one embodiment, one or more convolutional layers are used in conjunction with one or more transposed convolution operations to generate a large textured output image from a small input textured image while preserving global features and texture, according to various novel techniques described herein.

Abstract: The disclosure provides a learning framework that unifies both semantic segmentation and semantic edge detection. A learnable recurrent message passing layer is disclosed where semantic edges are considered as explicitly learned gating signals to refine segmentation and improve dense prediction quality by finding compact structures for message paths. The disclosure includes a method for coupled segmentation and edge learning. In one example, the method includes: (1) receiving an input image, (2) generating, from the input image, a semantic feature map, an affinity map, and a semantic edge map from a single backbone network of a convolutional neural network (CNN), and (3) producing a refined semantic feature map by smoothing pixels of the semantic feature map using spatial propagation, and controlling the smoothing using both affinity values from the affinity map and edge values from the semantic edge map.

Abstract: Apparatuses, systems, and techniques to train one or more neural networks. In at least one embodiment, one or more neural networks are trained to perform segmentation tasks based at least in part on training data comprising bounding box annotations.

Abstract: Apparatuses, systems, and techniques are presented to generate images. In at least one embodiment, one or more neural networks are used to generate one or more images using one or more pixel weights determined based, at least in part, on one or more sub-pixel offset values.

Abstract: Apparatuses, systems, and techniques are presented to generate images. In at least one embodiment, one or more neural networks are used to generate one or more images using one or more pixel weights determined based, at least in part, on one or more sub-pixel offset values.

Abstract: Apparatuses, systems, and techniques are presented to generate images. In at least one embodiment, one or more neural networks are used to generate one or more images using one or more pixel weights.

Abstract: Apparatuses, systems, and techniques are presented to generate images. In at least one embodiment, one or more neural networks are used to generate one or more images using one or more pixel weights determined based, at least in part, on one or more sub-pixel offset values.

Abstract: Inverse rendering estimates physical scene attributes (e.g., reflectance, geometry, and lighting) from image(s) and is used for gaming, virtual reality, augmented reality, and robotics. An inverse rendering network (IRN) receives a single input image of a 3D scene and generates the physical scene attributes for the image. The IRN is trained by using the estimated physical scene attributes generated by the IRN to reproduce the input image and updating parameters of the IRN to reduce differences between the reproduced input image and the input image. A direct renderer and a residual appearance renderer (RAR) reproduce the input image. The RAR predicts a residual image representing complex appearance effects of the real (not synthetic) image based on features extracted from the image and the reflectance and geometry properties. The residual image represents near-field illumination, cast shadows, inter-reflections, and realistic shading that are not provided by the direct renderer.

Abstract: Apparatuses, systems, and techniques to scale textured images using a Fourier transform in conjunction with one or more neural networks. In at least one embodiment, a neural network generates an expanded image from an input image by applying a Fourier transform to one or more feature maps generated by said neural network and up-scaling one or more resulting frequency domain feature maps before generating an expanded output image based on up-scaled feature maps.

Abstract: Apparatuses, systems, and techniques for texture synthesis from small input textures in images using convolutional neural networks. In at least one embodiment, one or more convolutional layers are used in conjunction with one or more transposed convolution operations to generate a large textured output image from a small input textured image while preserving global features and texture, according to various novel techniques described herein.

Abstract: Disclosed are an electronic cigarette and a vaporizer thereof which are suitable for adjusting air inflow with a single hand. The electronic cigarette vaporizer includes a vaporizer base and a liquid reservoir connected to the vaporizer. An atomizing cavity is formed by an internal cavity obtained by the connection of the liquid reservoir and the vaporizer base. An air inlet passage communicated with the atomizing cavity is arranged at the liquid reservoir. An adjusting knob at the entrance of the air inlet passage for adjusting air inflow is also arranged at the liquid reservoir. An air inlet is arranged at the adjusting knob. By rotating the adjusting knob, the air inflow can be adjusted by relative position changes of the air inlet and the entrance of the air inlet passage.

Abstract: Apparatuses, systems, and techniques to enhance video. In at least one embodiment, one or more neural networks are used to create, from a first video, a second video having a higher frame rate, higher resolution, or reduced number of missing or corrupt video frames.

Abstract: The present invention provides a method and apparatus for verifying images based on image verification codes, the method comprising selecting an identification image and multiple candidate images from an image gallery, where the candidate image comprise interference images and correct images corresponding to the identification image. The method also includes providing hint information for the identification image, the candidate images, and relationships between the identification image and the correct images. The method also includes receiving selection information of images selected from the candidate images. The method also includes determining if the verification passed when the correct images are determined to have been selected based on the selection information or, determining that verification has failed when the correct images are determined not to have been selected based on the selection information.

Abstract: Embodiments of the present disclosure provide a pixel unit, a display substrate, and a method for manufacturing the display substrate. The pixel unit includes: a first sub-pixel region; a second sub-pixel region located at a periphery of the first sub-pixel region and surrounding the first sub-pixel region; and a third sub-pixel region located at a periphery of the second sub-pixel region and surrounding the second sub-pixel region.