Abstract: One embodiment provides a method of using a computing device for image-to-image translation including accessing an image file containing a first amount of data. The computing device inputs the image file into a convolutional neural network (CNN). The CNN includes multiple Fourier layers. Each Fourier layer includes a Fourier transform, a linear feature transformation in a frequency domain and an inverse Fourier transform. Each linear feature transformation in the frequency domain is shared by different frequency components to reduce a number of parameters. The CNN outputs an output image file that includes contents that are translated from the input image file.

Abstract: A system includes a processor that executes computer executable components stored in a memory and a set of at least two disparate sound detection devices that receive sound waves at at least two receiving locations. The computer executable components include a determination component that determines amplitude, phase, and frequency of the respective received sound waves by the sound detection devices. The computer executable components include an analysis component that determines a phase difference between and a frequency spectrum of the received sound waves to determine a location of a potential a gas leak. The computer executable components include an identification component that identifies a location of a potential gas leak.

Abstract: An approach is disclosed that receives a time sequence of low-resolution images, each of the low-resolution images depicting a physics event. The approach interpolates two adjacent low-resolution images to a higher spatial-resolution interpolated image between a first and a second time. The approach then inputs a previous high-resolution image and the interpolated image to a neural network that includes a physics constraint that corresponds to the physics event. A new high-resolution image is received from the neural network, with the new image corresponding to the inputted previous high-resolution image and the interpolated image. The neural network is trained to minimize the mean squared difference between a smoothed version of the output and the input at the same time plus the difference between the high-resolution output and a second relevant physics equation.

Abstract: Imaging data is received from one or more imaging sources by a computing device over one or more networks. The imaging data is for a selected spatiotemporal dimension and indicates an object detected by at least one imaging source of the one or more imaging sources. Object-specific data for a selected type of object and for the selected spatiotemporal dimension is obtained from at least one data source. The imaging data and the object-specific data for the selected spatiotemporal dimension are processed to identify whether the object detected by the at least one imaging source is misclassified as the selected type of object. The processing employs an artificial intelligence model trained to analyze the imaging data and the object-specific data. Based on the processing, an indication of whether the object is misclassified is output.

Abstract: An embodiment includes identifying a tree type of vegetation depicted in an image. The embodiment segments that portion of the image using edge-detection processing resulting in a contour line that defines a tree perimeter. The embodiment detects that the tree is within a buffer distance from a power line. The embodiment determines the tree's species by comparing the contour line to candidate contour lines of different tree species and calculates a diameter of the tree's crown using the contour line. The embodiment estimates the tree's height using the species and the diameter of the crown. The embodiment calculates a risk value for the tree based on a risk of contact between the power line and the tree and issues a work order to maintain the tree to prevent contact with the power line.

Abstract: In an approach to generating super resolution images, a computer selects a latent vector associated with a high resolution image from a plurality of latent vectors of a generative neural network model. A computer generates a super resolution image from the selected latent vector. A computer downscales the super resolution image to match a size of a plurality of low resolution images. A computer computes a difference between the super resolution image and each of the plurality of low resolution images. A computer determines a minimum difference of the difference between the super resolution image and each of the plurality of low resolution images. A computer determines the minimum difference meets a stopping criteria. A computer transmits the super resolution image to a user. A computer stores the super resolution image.

Abstract: According to one embodiment, a method, computer system, and computer program product for dynamic image reconstruction is provided. The present invention may include training a physics-informed neural network (PINN) using received training data; receiving wind field velocity measurements in a geographical domain from one or more weather information sources and air pollutant concentration measurements of one or more air pollutants from the geographical domain using a sparse sensor network to produce a plurality of input data; processing the plurality of input data through multiple physics-informed masked convolutional layers in the trained PINN to perform a three-dimensional partial convolution process; and performing the dynamic image reconstruction on the processed plurality of input data using the trained PINN to generate a reconstructed continuous spatial map of the geographical domain.

Abstract: One embodiment provides a method of using a computing device for image-to-image translation including accessing an image file containing a first amount of data. The computing device inputs the image file into a convolutional neural network (CNN). The CNN includes multiple Fourier layers. Each Fourier layer includes a Fourier transform, a linear feature transformation in a frequency domain and an inverse Fourier transform. Each linear feature transformation in the frequency domain is shared by different frequency components to reduce a number of parameters. The CNN outputs an output image file that includes contents that are translated from the input image file.

Abstract: An approach is disclosed that receives a time sequence of low-resolution images, each of the low-resolution images depicting a physics event. The approach interpolates two adjacent low-resolution images to a higher spatial-resolution interpolated image between a first and a second time. The approach then inputs a previous high-resolution image and the interpolated image to a neural network that includes a physics constraint that corresponds to the physics event. A new high-resolution image is received from the neural network, with the new image corresponding to the inputted previous high-resolution image and the interpolated image. The neural network is trained to minimize the mean squared difference between a smoothed version of the output and the input at the same time plus the difference between the high-resolution output and a second relevant physics equation.

Abstract: Road condition prediction for potentially hazardous road segments is described. For roadways that may contain accumulated frozen precipitation, a road segment for road condition prediction is selected based on weather conditions. Various models including a solar radiation budget model, a permanent structures model, a dynamic structures model, and a road condition model are generated for the selected road segment and account for shading effects on the road segment caused by objects near the road segment. A road condition prediction for hazardous conditions on the road segment is determined based on the road condition model and provided to a driver to alert the driver of any potentially hazardous conditions.

Abstract: A method, computer program product and system to generate higher resolution geospatial images is provided. A processor receives time sequenced spatial data images at a first resolution. A processor determines from the plurality of spatial data images physics laws applicable to the spatial data images. A processor subdivides each of the plurality of spatial data images into a plurality of small spatial region images. A processor solves each of the physics laws in each of the small spatial region images. A processor trains a neural network to apply each of the physics laws to each small spatial region image by applying a regional physics law loss function. A processor determines the most applicable regional physics law based on the difference between the small spatial region image and the image predicted for that region by the physics law. A processor generates a second higher-resolution image than the first resolution.

Abstract: In an approach for estimating emission source location from satellite plume data, a processor creates a dataset of plume concentration data. A processor down samples the dataset to an array at satellite resolution. A processor partitions the array into two separate datasets according to a preset proportion. A processor trains two machine learning models on at least one of the two separate datasets, wherein a first machine learning model of the two machine learning models is for identifying a presence of a plume and a second machine learning model of the two machine learning models is for identifying a source position and magnitude of the plume. A processor applies the two machine learning models to new concentration data.

Abstract: An embodiment includes identifying a tree type of vegetation depicted in an image. The embodiment segments that portion of the image using edge-detection processing resulting in a contour line that defines a tree perimeter. The embodiment detects that the tree is within a buffer distance from a power line. The embodiment determines the tree's species by comparing the contour line to candidate contour lines of different tree species and calculates a diameter of the tree's crown using the contour line. The embodiment estimates the tree's height using the species and the diameter of the crown. The embodiment calculates a risk value for the tree based on a risk of contact between the power line and the tree and issues a work order to maintain the tree to prevent contact with the power line.

Abstract: In some examples, a method of vector-raster data fusion includes receiving vector data for a geographical location, and statistically analyzing the vector data to obtain vector statistics. In some examples the method further includes rasterizing the vector statistics, and storing at least one of the vector data and the rasterized vector statistics together in a key-value store together with previously stored raster data for the geographical location. In some examples, the vector data further includes metadata, and the method further includes storing the metadata in at least one of the key-value store or a separate vector database.

Abstract: Aspects of the invention include includes detecting, using a first machine learning model, a first well pad at a first location based at least in part on a first set of data comprising spectral data describing a gas emission from the first location. Detecting an environmental event within a threshold distance of the well pad. Determining a probability of damage to the first well pad from the environmental event.

Abstract: Aspects of the invention include using a controller to control a transceiver to transmit a sensor query signal to a first sensor at a first location of one or more locations having one or more sensors, wherein the sensor query signal energizes a first power supply for the first sensor, wherein energizing the power supply causes the first sensor to perform a sensor reading at the first location and transmit to the transceiver an encoded response signal representing the sensor reading, and analyzing, using the controller, the encoded response signal to determine the sensor reading at the first location.

Abstract: Methods and systems for managing vegetation include training a machine learning model based on an image of a training data region before a weather event, an image of the training data region after the weather event, and information regarding the weather event. A risk score is generated for a second region using the trained machine learning model based on an image of the second region and predicted weather information for the second region. The risk score is determined to indicate high-risk vegetation in the second region. A corrective action is performed to reduce the risk of vegetation in the second region.

Abstract: A system and method provide remote tracking of progress at construction sites. When possible, the progress can be monitored remotely via satellite imagery. But once satellite imagery cannot determine the progress that needs to be monitored at the construction site, multiple unmanned aerial vehicles (UAVs) that use RF emitters and receivers are deployed to the construction site to determine progress based on RF scans. A number and type of UAVs, along with their respective flight plans, are determined from site specifications and expected progress at the construction site. The UAVs are programmed with their respective flight plans, synchronized, then deployed to inspect the construction site using RF scans. The UAVs execute their respective flight plans, with some emitting RF signals and others receiving those emitted RF signals. The UAV RF data is then analyzed and compared to the site specifications to determine the progress at the construction site.

Abstract: A computer-implemented method executed by one or more satellites for assessing crop development by using synthetic aperture radar (SAR) is presented. The method includes generating SAR images from scanning fields including crops, monitoring grown of the crops within the fields during a predetermined time period, and estimating a height of the crops during the predetermined time period by using interferometric information from one or more of the SAR images and tracking change in height and growth rates. The method further includes differentiating between crops in different fields by monitoring changes in the height of the crops during an entire growing season.

Abstract: A rover includes a base having wheels and a propulsion system coupled to the wheels to propel the rover around a field. A tower is coupled to the base and extends over the base. Sensors are set on the tower and the base and are configured to sense environmental conditions around the rover at different elevations. A computing system includes a processor and memory. The memory is configured to receive measured data from the sensors and determine an amount and manner of water to be dispensed on plant life in the field.