Abstract: The exemplary embodiments relate to converting Standard Dynamic Range (SDR) content to High Dynamic Range (HDR) content using a machine learning system. In some embodiments, the neural network is trained to convert an input SDR image into an HDR image using the encoded representation of a training SDR image and a training HDR image. In other embodiments, the neural network is trained to convert an input SDR image into an HDR image using a predefined set of color grading actions and the training images.

Abstract: The exemplary embodiments relate to converting Standard Dynamic Range (SDR) content to High Dynamic Range (HDR) content using a machine learning system. In some embodiments, the neural network is trained to convert an input SDR image into an HDR image using the encoded representation of a training SDR image and a training HDR image. In other embodiments, the neural network is trained to convert an input SDR image into an HDR image using a predefined set of color grading actions and the training images.

Abstract: The exemplary embodiments relate to converting Standard Dynamic Range (SDR) content to High Dynamic Range (HDR) content using a machine learning system. In some embodiments, the neural network is trained to convert an input SDR image into an HDR image using the encoded representation of a training SDR image and a training HDR image. In other embodiments, the neural network is trained to convert an input SDR image into an HDR image using a predefined set of color grading actions and the training images.

Abstract: A process receives, with a processor, a plurality of image device properties corresponding to an image capture device. Further the process selects, with the processor, one or more of the plurality of image device properties based on one or more pipeline criteria. The pipeline criteria indicate one or more weights corresponding to the plurality of image device properties. Additionally, the process truncates, with the processor, the one or more selected image device properties to a predetermined number of characters based on a human readability factor. The human readability factor is provided by the one or more pipeline criteria. The process automatically generates, with the processor, a device parameter data structure by ordering the truncated one or more image device properties according to the one or more weights. Moreover, the process transmits, with the processor, the device parameter data structure through a video post-production pipeline.

Abstract: A process receives, with a processor, a plurality of image device properties corresponding to an image capture device. Further the process selects, with the processor, one or more of the plurality of image device properties based on one or more pipeline criteria. The pipeline criteria indicate one or more weights corresponding to the plurality of image device properties. Additionally, the process truncates, with the processor, the one or more selected image device properties to a predetermined number of characters based on a human readability factor. The human readability factor is provided by the one or more pipeline criteria. The process automatically generates, with the processor, a device parameter data structure by ordering the truncated one or more image device properties according to the one or more weights. Moreover, the process transmits, with the processor, the device parameter data structure through a video post-production pipeline.

Abstract: This invention uses a novel synthesis of several different compression methods, some original and some known in the art, to achieve compression of digitized video image sequences. Images are color-reduced by a factor of two in both directions using a colorspace that requires minimal computation for reconstruction of RGB pixel values. Images are then transformed into a `token sequence` corresponding to a series of 4.times.4 blocks, using one of seven transformation methods, some dependent on previous frames and some dependent only on pixel values in neighboring blocks. The resulting token sequence is then compressed using known lossless methods, including Huffman coding.