Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing deep learning to intelligently determine compression settings for compressing a digital image. For instance, the disclosed system utilizes a neural network to generate predicted perceptual quality values for compression settings on a compression quality scale. The disclosed system fits the predicted compression distortions to a perceptual distortion characteristic curve for interpolating predicted perceptual quality values across the compression settings on the compression quality scale. Additionally, the disclosed system then performs a search over the predicted perceptual quality values for the compression settings along the compression quality scale to select a compression setting based on a perceptual quality threshold. The disclosed system generates a compressed digital image according to compression parameters for the selected compression setting.

Abstract: A system for allocating resources across equipment that operate to serve one or more loads of a building. The system includes one or more memory devices storing instructions that cause one or more processors to receive operational data defining at least one of planned loads to be served by the equipment or utility rates for one or more time steps within a simulation period, determine whether the operational data define the planned loads or the utility rates for each time step within the simulation period, and in response to a determination that the operational data do not define the planned loads or the utility rates for each time step within the simulation period, identify one or more time steps for which the planned loads or the utility rates are not defined and initiate an action to define the planned loads or the utility rates for the identified time steps.

Abstract: Disclosed herein are systems and methods for virtual mapping in autonomous vehicle operation. The systems and methods include navigating, by a computing system, a virtual model of an autonomous vehicle through a virtual environment corresponding to an interior space of a real-world building; generating, by the computing system, a virtual map of the interior space for the autonomous vehicle based on the navigation through the virtual environment; and transmitting, by a communication device of the computing system, the virtual map to the autonomous vehicle for navigating the interior space.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing deep learning to intelligently determine compression settings for compressing a digital image. For instance, the disclosed system utilizes a neural network to generate predicted perceptual quality values for compression settings on a compression quality scale. The disclosed system fits the predicted compression distortions to a perceptual distortion characteristic curve for interpolating predicted perceptual quality values across the compression settings on the compression quality scale. Additionally, the disclosed system then performs a search over the predicted perceptual quality values for the compression settings along the compression quality scale to select a compression setting based on a perceptual quality threshold. The disclosed system generates a compressed digital image according to compression parameters for the selected compression setting.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing deep learning to intelligently determine compression settings for compressing a digital image. For instance, the disclosed system utilizes a neural network to generate predicted perceptual quality values for compression settings on a compression quality scale. The disclosed system fits the predicted compression distortions to a perceptual distortion characteristic curve for interpolating predicted perceptual quality values across the compression settings on the compression quality scale. Additionally, the disclosed system then performs a search over the predicted perceptual quality values for the compression settings along the compression quality scale to select a compression setting based on a perceptual quality threshold. The disclosed system generates a compressed digital image according to compression parameters for the selected compression setting.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing deep learning to intelligently determine compression settings for compressing a digital image. For instance, the disclosed system utilizes a neural network to generate predicted perceptual quality values for compression settings on a compression quality scale. The disclosed system fits the predicted compression distortions to a perceptual distortion characteristic curve for interpolating predicted perceptual quality values across the compression settings on the compression quality scale. Additionally, the disclosed system then performs a search over the predicted perceptual quality values for the compression settings along the compression quality scale to select a compression setting based on a perceptual quality threshold. The disclosed system generates a compressed digital image according to compression parameters for the selected compression setting.

Abstract: A method includes operating equipment to consume energy resources including energy or power purchased from a utility, and obtaining a block-and-index rate profile for a future time period. The block-and-index rate profile includes a block rate and a block size for each of a plurality of sub-periods in the future time period. The block size for a sub-period identifies an amount of energy or power priced at the block rate for the sub-period. The method also includes applying the block-and-index rate profile in an optimization process for the equipment over the time period, running the optimization process, and allocating energy resources to the equipment over the time period in accordance with a result of the optimization process.

Abstract: A method includes operating equipment to consume energy resources including energy or power purchased from a utility, and obtaining a block-and-index rate profile for a future time period. The block-and-index rate profile includes a block rate and a block size for each of a plurality of sub-periods in the future time period. The block size for a sub-period identifies an amount of energy or power priced at the block rate for the sub-period. The method also includes applying the block-and-index rate profile in an optimization process for the equipment over the time period, running the optimization process, and allocating energy resources to the equipment over the time period in accordance with a result of the optimization process.

Abstract: A system for allocating resources across equipment that operate to serve one or more loads of a building. The system includes one or more memory devices storing instructions that cause one or more processors to receive operational data defining at least one of planned loads to be served by the equipment or utility rates for one or more time steps within a simulation period, determine whether the operational data define the planned loads or the utility rates for each time step within the simulation period, and in response to a determination that the operational data do not define the planned loads or the utility rates for each time step within the simulation period, identify one or more time steps for which the planned loads or the utility rates are not defined and initiate an action to define the planned loads or the utility rates for the identified time steps.