Abstract: Methods, systems, and apparatus for obtaining input features representative of a region of space, processing an input comprising the input features through the ML model to generate a prediction describing predicted features of the region of space, obtaining result features describing the region of space, determining a value of at least one evaluation metric that relates the predicted features and the result features, that at least one evaluation metric including one of a distance score, a pyramiding density error, and min-max intersection over union (IOU) score, and training the ML model responsive to the at least one evaluation metric. Other implementations of this aspect include corresponding systems, apparatus, and computer programs, configured to perform the actions of the methods, encoded on computer storage devices.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a synthesized signal. A computer-implemented system obtains generator input data including an input signal having one or more first characteristics, processes the generator input data to generate output data including a synthesized signal having one or more second characteristics using a generator neural network, and outputs the synthesized signal to a device. The generator neural network is trained, based on a plurality of training examples, with a discriminator neural network. The discriminator neural network is configured to process discriminator input data that combines a discriminator input signal having the one or more second characteristics with at least a portion of generator input data to generate a prediction of whether the discriminator input signal is a real signal or a synthesized signal.

Abstract: Methods, systems, and apparatus for providing a ML model for inference, the ML model having been trained using a first set of training data to provide predictions associated with an adverse event, after training of the ML model, receiving data from one or more data sources, the data representative of characteristics relevant to predictions associated with the adverse event, providing a second set of training data, determining, by a trigger module, a trigger decision based on a set of signals at least partially determined from the second set of training data, the trigger decision indicating whether the ML model is to be one of updated and retrained based on the second set of training data, and selectively executing one of updating and retraining of the ML model using at least a portion of the second set of training data in response to the trigger decision.

Abstract: Methods, systems, and apparatus for obtaining a first plurality of data elements, each data element representing a fire-related metric of a geographic region, determining, using at least a subset of the first data elements, one or more values representing one or more derived fire-related metrics, associating the one or more values with the first data elements, obtaining a second plurality of data elements, each data element representing a fire-related metric of the geographic region, and training a machine learning (ML) model using at least a subset of the first plurality of data elements, at least a subset of the second plurality of data elements, and values associated with the subset of the first plurality of data elements to provide a trained ML model.

Abstract: Methods, systems, and apparatus for receiving a request for a risk assessment for a parcel, receiving a set of images for the parcel, the set of images including two or more images, each image having an image scale and an image resolution that is different from other images in the set of images, providing a first-level feature embedding and a second-level feature embedding, the first-level feature embedding being provided by processing a first-level image through a first-level machine learning (ML) model, and the second-level feature embedding being provided by processing a second-level image through a second-level ML model, determining a risk assessment at least partially by processing each of the first-level feature embedding and a second-level feature embedding through a fusion network, and providing a representation of the risk assessment for display.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a synthesized signal. In some implementations, a computer-implemented system obtains generator input data including at least an input signal having one or more first characteristics, processes the generator input data to generate output data including a synthesized signal having one or more second characteristics using a generator neural network, and outputs the synthesized signal to a device. The generator neural network is trained, based on a plurality of training examples, with a discriminator neural network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a synthesized signal. In some implementations, a computer-implemented system obtains generator input data including at least an input signal having one or more first characteristics, processes the generator input data to generate output data including a synthesized signal having one or more second characteristics using a generator neural network, and outputs the synthesized signal to a device. The generator neural network is trained, based on a plurality of training examples, with a discriminator neural network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for identifying wildfire in satellite imagery. In some implementations, a server obtains a satellite image of a geographic region and a date corresponding to when the satellite image was generated. The server determines a number of pixels in the satellite image that are indicated as on fire. The server obtains satellite imagery of the geographic region from before the date. The server generates a statistical distribution from the satellite imagery. The server determines a likelihood that the satellite image illustrates fire based on a comparison of the determined number of pixels in the satellite image that are indicated as on fire to the generated statistical distribution. The server can compare the determined likelihood to a threshold. In response to comparing the determined likelihood to the threshold, the server provides an indication that the satellite image illustrates fire.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a temporal range of a fire. In some implementations, a server obtains a date when a fire occurred within a region. The server obtains satellite imagery of the region from before the date when the fire occurred. The server generates a first statistical distribution from the satellite imagery. The server determines a start date of the fire using the first statistical distribution. The server obtains second satellite imagery of the region from before and after the start date. The server selects a second set of imagery from the second satellite imagery from before the start date. The server generates a second statistical distribution from the second set of imagery. The server determines an end date of the fire using the second statistical distribution. The server provides the start date and the end date for output.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating high-resolution fire distribution maps. In some implementations, a computer-implemented system obtains a low-resolution distribution map indicating fire distribution of an area with fire burning and a reference map indicating features of the same area. The system processes the low-resolution distribution map and the reference map using a generator neural network to generate output data including a high-resolution synthesized distribution map indicating fire distribution of the area. The generator neural network is trained, based on a plurality of training examples, with a discriminator neural network that outputs a prediction of whether an input to the discriminator neural network is a real distribution map or a synthesized distribution map.