Abstract: Disclosed is a method of analyzing an energy storage device including normalizing charge and discharge voltage profiles to obtain normalized charge and discharge voltage profiles as a function of a normalized capacity; averaging the normalized voltage profiles to obtain an averaged voltage profile as a function of the normalized capacity; determining an internal resistance between the charge or discharge voltage profiles and the normalized averaged voltage profile; determining a current-resistance (IR) potential responsive to a measured current and the determined internal resistance; determining an IR-free voltage charge or discharge profile of the energy storage device; and determining a ratio between a capacity change and a voltage potential change in a charging or discharging process of the energy storage device over a predetermined interval of time responsive to the determined IR-free voltage charge or discharge profiles.

Abstract: Energy storage cell qualification and related systems, methods, and devices are disclosed. A method of qualifying rechargeable battery cells includes taking measurements on the rechargeable battery cells, determining specific capacity distributions of the rechargeable battery cells as a function of a number of discharge cycles based on the measurements, determining one or more specific capacity thresholds to separate the specific capacity distributions of the rechargeable battery cells into two or more classifications, and qualifying the rechargeable battery cells into the two or more classifications based, at least in part, on the specific capacity distributions and the one or more specific capacity thresholds. A method of implementing rechargeable battery cells into product manufacturing and qualifying the rechargeable battery cells, and deploying those of the rechargeable battery cells qualified into a first classification of the two or more classifications into the product.

Abstract: A method of operating an energy storage device comprises applying external pressure onto an electrochemical cell during charging and/or discharging. The cell comprises a sulfur-based cathode within a compressible vessel. An energy storage system may include plates and at least one electrochemical cell that includes—within a compressible vessel—a cathode, an anode, and an electrolyte. The cathode comprises a sulfur-based active material. In a method of assembling a system for operating an energy storage device, at least one electrochemical cell is disposed between a pair of plates and adjacent pressure sensor(s). A distance between the plates is fixed at which a measured pressure is within a range from greater than about 0 kPa to about 689 kPa above ambient pressure. The plate-separation distance is maintained while charging or discharging the electrochemical cell(s).

Abstract: Energy storage cell qualification and related systems, methods, and devices are disclosed. A method of qualifying rechargeable battery cells includes taking measurements on the rechargeable battery cells, determining specific capacity distributions of the rechargeable battery cells as a function of a number of discharge cycles based on the measurements, determining one or more specific capacity thresholds to separate the specific capacity distributions of the rechargeable battery cells into two or more classifications, and qualifying the rechargeable battery cells into the two or more classifications based, at least in part, on the specific capacity distributions and the one or more specific capacity thresholds. A method of implementing rechargeable battery cells into product manufacturing and qualifying the rechargeable battery cells, and deploying those of the rechargeable battery cells qualified into a first classification of the two or more classifications into the product.

Abstract: Systems and methods for texture hallucination with a large upscaling factor are described. Embodiments of the systems and methods may receive an input image and a reference image, extract an upscaled feature map from the input image, match the input image to a portion of the reference image, wherein a resolution of the reference image is higher than a resolution of the input image, concatenate the upscaled feature map with a reference feature map corresponding to the portion of the reference image to produce a concatenated feature map, and generate a reconstructed image based on the concatenated feature map using a machine learning model trained with a texture loss and a degradation loss, wherein the texture loss is based on a high frequency band filter, and the degradation loss is based on a downscaled version of the reconstructed image.

Abstract: A style of a digital image is transferred to another digital image of arbitrary resolution. A high-resolution (HR) content image is segmented into several low-resolution (LR) patches. The resolution of a style image is matched to have the same resolution as the LR content image patches. Style transfer is then performed on a patch-by-patch basis using, for example, a pair of feature transforms—whitening and coloring. The patch-by-patch style transfer process is then repeated at several increasing resolutions, or scale levels, of both the content and style images. The results of the style transfer at each scale level are incorporated into successive scale levels up to and including the original HR scale. As a result, style transfer can be performed with images having arbitrary resolutions to produce visually pleasing results with good spatial consistency.

Abstract: Systems and methods for texture hallucination with a large upscaling factor are described. Embodiments of the systems and methods may receive an input image and a reference image, extract an upscaled feature map from the input image, match the input image to a portion of the reference image, wherein a resolution of the reference image is higher than a resolution of the input image, concatenate the upscaled feature map with a reference feature map corresponding to the portion of the reference image to produce a concatenated feature map, and generate a reconstructed image based on the concatenated feature map using a machine learning model trained with a texture loss and a degradation loss, wherein the texture loss is based on a high frequency band filter, and the degradation loss is based on a downscaled version of the reconstructed image.

Abstract: The present disclosure relates to polar decoding methods and apparatus. One example method includes obtaining a polar code sequence, determining at least one first information bit length corresponding to a code length of the polar code sequence, and decoding the polar code sequence based on blind detection by using the code length and the at least one first information bit length. The at least one first information bit length is a part of a preset information bit length set corresponding to the code length. A remaining part of the preset information bit length set is at least one second information bit length. The at least one second information bit length is excluded from being used for the blind detection. Each second information bit length is less than one or more information bit lengths in the at least one first information bit length.

Abstract: In implementations of transferring image style to content of a digital image, an image editing system includes an encoder that extracts features from a content image and features from a style image. A whitening and color transform generates coarse features from the content and style features extracted by the encoder for one pass of encoding and decoding. Hence, the processing delay and memory requirements are low. A feature transfer module iteratively transfers style features to the coarse feature map and generates a fine feature map. The image editing system fuses the fine features with the coarse features, and a decoder generates an output image with content of the content image in a style of the style image from the fused features. Accordingly, the image editing system efficiently transfers an image style to image content in real-time, without undesirable artifacts in the output image.

Abstract: A style of a digital image is transferred to another digital image of arbitrary resolution. A high-resolution (HR) content image is segmented into several low-resolution (LR) patches. The resolution of a style image is matched to have the same resolution as the LR content image patches. Style transfer is then performed on a patch-by-patch basis using, for example, a pair of feature transforms—whitening and coloring. The patch-by-patch style transfer process is then repeated at several increasing resolutions, or scale levels, of both the content and style images. The results of the style transfer at each scale level are incorporated into successive scale levels up to and including the original HR scale. As a result, style transfer can be performed with images having arbitrary resolutions to produce visually pleasing results with good spatial consistency.

Abstract: A style of a digital image is transferred to another digital image of arbitrary resolution. A high-resolution (HR) content image is segmented into several low-resolution (LR) patches. The resolution of a style image is matched to have the same resolution as the LR content image patches. Style transfer is then performed on a patch-by-patch basis using, for example, a pair of feature transforms—whitening and coloring. The patch-by-patch style transfer process is then repeated at several increasing resolutions, or scale levels, of both the content and style images. The results of the style transfer at each scale level are incorporated into successive scale levels up to and including the original HR scale. As a result, style transfer can be performed with images having arbitrary resolutions to produce visually pleasing results with good spatial consistency.

Abstract: A style of a digital image is transferred to another digital image of arbitrary resolution. A high-resolution (HR) content image is segmented into several low-resolution (LR) patches. The resolution of a style image is matched to have the same resolution as the LR content image patches. Style transfer is then performed on a patch-by-patch basis using, for example, a pair of feature transforms—whitening and coloring. The patch-by-patch style transfer process is then repeated at several increasing resolutions, or scale levels, of both the content and style images. The results of the style transfer at each scale level are incorporated into successive scale levels up to and including the original HR scale. As a result, style transfer can be performed with images having arbitrary resolutions to produce visually pleasing results with good spatial consistency.

Abstract: In implementations of transferring image style to content of a digital image, an image editing system includes an encoder that extracts features from a content image and features from a style image. A whitening and color transform generates coarse features from the content and style features extracted by the encoder for one pass of encoding and decoding. Hence, the processing delay and memory requirements are low. A feature transfer module iteratively transfers style features to the coarse feature map and generates a fine feature map. The image editing system fuses the fine features with the coarse features, and a decoder generates an output image with content of the content image in a style of the style image from the fused features. Accordingly, the image editing system efficiently transfers an image style to image content in real-time, without undesirable artifacts in the output image.

Abstract: The present disclosure relates to polar decoding methods and apparatus. One example method includes obtaining a polar code sequence, determining at least one first information bit length corresponding to a code length of the polar code sequence, and decoding the polar code sequence based on blind detection by using the code length and the at least one first information bit length. The at least one first information bit length is a part of a preset information bit length set corresponding to the code length. A remaining part of the preset information bit length set is at least one second information bit length. The at least one second information bit length is excluded from being used for the blind detection. Each second information bit length is less than one or more information bit lengths in the at least one first information bit length.