Abstract: Systems and methods for detection and localization of image and document forgery. The method can include the step of receiving a dataset having a plurality of authentic images and a plurality of manipulated images. The method can also include the step of benchmarking a plurality of image forgery algorithms using the dataset. The method can further include the step of generating a plurality of receiver operating characteristic (ROC) curves for each of the plurality of image forgery algorithms. The method also includes the step of calculating a plurality of area under curve metrics for each of the plurality of ROC curves. The method further includes the step of training a neural network for image forgery based on the plurality of area under curve metrics.

Abstract: Systems and methods for roof area and slope estimation using a point set are provided. The system selects roof structure points having a high probability of being positioned on a top surface of a structure present in the region of interest point set. Then, the system determines a footprint of the structure associated with the selected roof structure points. The system determines a distribution of the slopes of the roof structure points and generates a slope distribution report indicative of prominent slopes of the roof structure and each slope's contribution toward (percentage composition of) the total roof structure. The system then determines an area of the roof structure based on the footprint of the structure and the slope distribution report.

Abstract: A system and method for generating a parametric model of a roof structure comprising a processor in communication with a memory. The system receives a plurality of parameters of each roof component composing the roof structure and performs a geometry creation based on the received plurality of parameters. The system generates a constrained three-dimensional geometry based on an output of the geometry creation, and displays a three-dimensional model of the roof structure based on the constrained three-dimensional geometry.

Abstract: Computer vision systems and methods for localizing image forgery are provided. The system generates a constrained convolution via a plurality of learned rich filters. The system trains a convolutional neural network with the constrained convolution and a plurality of images of a dataset to learn a low level representation of each image among the plurality of images. The low level representation is indicative of a statistical signature of at least one source camera model of each image. The system can determine a splicing manipulation localization by the trained convolutional neural network.

Abstract: Systems and methods for adjusting the location and scale of a three-dimensional (“3D”) model of an object to conform to a georeferenced point cloud are provided herein. The 3D model and the georeferenced point cloud are rendered in a shared 3D coordinate system and the 3D model and the georeferenced point cloud are aligned from a first point of view. An affine transformation matrix is calculated based on a best fitting plane of the point cloud and a corresponding face of the 3D model to align the best fitting plane and the corresponding face of the 3D model. The affine transformation matrix is then applied to all coordinates of the 3D model to generate a new 3D model that aligns with the georeferenced point cloud from a second point of view.

Abstract: Computer vision systems and methods for determining roof conditions from imagery using segmentation networks are provided. The system obtains at least one image from an image database having a roof structure present therein, and determines a footprint of the roof structure using a neural network. Based on segmentation processing by the neural network, the system generates a single channel image that maps each pixel in the at least one image to a binary classification indicative of whether each pixel is or is not representative of a roof structure and executes a contour extraction algorithm on the single channel image to determine the footprint of the roof structure.

Abstract: Machine learning systems and methods for interactive concept searching using attention scoring are provided. The system receives textual data. The system identifies one or more word representations of the textual data. The system further receives a concept. The system determines a score indicative of a likelihood of each of the one or more word representations being representative of the concept using an attention scoring process having a temperature variable. The system generates a dataset for training and evaluation based at least in part on the score. The dataset includes the one or more word representations and concept. The system further processes the dataset to train one or more deep active learning models capable of the interactive concept search.

Abstract: A method of evaluating the driving behavior in a vehicle. The method includes determining values of a plurality of parameters of the operation of a first vehicle in a first road segment, determining values of the plurality of parameters for one or more second vehicles in a second road segment having similar properties to those of the first road segment, comparing the determined values of the first vehicle and the one or more second vehicles and providing an evaluation of the driving behavior of the first vehicle, responsive to the comparison.

Abstract: Computer vision systems and methods for high-fidelity representation of complex 3D surfaces using deep unsigned distance embeddings are provided. The system receives data associated with the 3D surface. The system processes the data based at least in part on one or more computer vision models to predict an unsigned distance field and a normal vector field. The unsigned distance field is indicative of proximity to the 3D surface and includes a predicted closest unsigned distance to a surface point of the 3D surface from a given point in a 3D space. The normal vector field is indicative of a surface orientation of the 3D surface and includes a predicted normal vector to the surface point closest to the given point. The system further determines the 3D surface representation based at least in part on the unsigned distance field and the normal vector field.

Abstract: Machine learning (ML) systems and methods for fact extraction and claim verification are provided. The system receives a claim and retrieves a document from a dataset. The document has a first relatedness score higher than a first threshold, which indicates that ML models of the system determine that the document is most likely to be relevant to the claim. The dataset includes supporting documents and claims including a first group of claims supported by facts from more than two supporting documents and a second group of claims not supported by the supporting documents. The system selects a set of sentences from the document. The set of sentences have second relatedness scores higher than a second threshold, which indicate that the ML models determine that the set of sentences are most likely to be relevant to the claim. The system determines whether the claim includes facts from the set of sentences.

Abstract: Systems and methods for detecting and aligning land property boundaries on aerial imagery are provided. The system receives an aerial imagery having land properties. The system applies a feature encoder having a plurality of levels to the aerial imagery. A first level of the plurality of levels includes a convolution block and a discrete wavelet transform layer. The discrete wavelet transform layer decomposes an input feature tensor to the first level into a low-frequency band and a high-frequency band. The high-frequency band is cached and processed with side-convolutional blocks before the high-frequency band are passed to a feature decoder. The system applies the feature decoder to an output of the feature encoder based at least in part on one of inverse discrete wavelet transform layers. The system determines boundaries of the one or more land properties based at least in part on a boundary cross-entropy loss function.

Abstract: Systems and methods for lossless compression of tabular numeric data are provided. The system can include one or more data compression servers executing data compression system code to compress the tabular numeric data, a storage database to store the compressed tabular numeric data, and one or more data decompression servers to decompress the tabular numeric data for use. The one or more data compression servers, the storage database, and the one or more data decompression servers can communicate via a communication network. The system can receive the uncompressed tabular numeric data from one or more data generation systems, processes the uncompressed tabular numeric data with the data compression system code, and generate a compressed table of numeric information, which can be stored in the database, or later decompressed by the one or more data decompression servers.

Abstract: Systems and methods for generating augmented reality environments from 2D drawings are provided. The system performs a camera calibration process to determine how a camera transforms images from the real world into a 2D image plane. The system calculates a camera pose and determines an object position and an object orientation relative to a known coordinate system. The system detects and processes a 2D drawing/illustration and generates a 3D model from the 2D drawing/illustration. The system performs a rendering process, wherein the system generates an augmented reality environment which includes the 3D model superimposed on an image of the 2D drawing/illustration. The system can generate the augmented reality environment in real time, allowing the system to provide immediate feedback to the user. The images processed by the system can be from a video, from multiple image photography, etc.

Abstract: A system for modeling a roof structure comprising an aerial imagery database and a processor in communication with the aerial imagery database. The aerial imagery database stores a plurality of stereoscopic image pairs and the processor selects at least one stereoscopic image pair among the plurality of stereoscopic image pairs and related metadata from the aerial imagery database based on a geospatial region of interest. The processor identifies a target image and a reference image from the at least one stereoscopic pair and calculates a disparity value for each pixel of the identified target image to generate a disparity map. The processor generates a three dimensional point cloud based on the disparity map, the identified target image and the identified reference image. The processor optionally generates a texture map indicative of a three-dimensional representation of the roof structure based on the generated three dimensional point cloud.

Abstract: Systems and methods for generating property data packages from light detection and ranging (“LIDAR”) point clouds are provided. Data from the LIDAR systems is stored in a LIDAR database in communication with a computer processor that executes computer vision system code including a 3D point cloud model generator, a measurement tool(s) subsystem, a 3D wireframe model generator, and a measurement extraction subsystem. The system receives an indication of a region of interest (“ROI”) from a user, retrieves point cloud data associated with the ROI from the LIDAR database, generates a 3D point cloud model based on the point cloud data, generates a 3D wireframe model based on the 3D point cloud model and operator input, extracts measurement information from the 3D wireframe model, and transmits a data package to the user. The data package can include, but is not limited to, the 3D point cloud model, the 3D wireframe model, and the extracted measurement data.

Abstract: Systems and methods for fine adjustment of computerized roof models are provided. The system generates a 3D roof structure model based on at least one image obtained from an aerial imagery database. Alternatively, the system could retrieve at least one stored 3D roof structure model from a 3D roof structure model database. The system weighs (e.g., scores) each 3D roof structure model candidate and determines an optimal 3D roof structure model by applying a variable neighborhood search to a 3D roof structure model candidate having a highest confidence score among the weighed 3D roof structure model candidates.

Abstract: Systems and methods for detecting, extracting, and categorizing structure data from aerial imagery following a major weather event are provided. The system processes digital images and weather data to automatically detect, extract, and categorize structure data following a major weather event. After receiving an indication of a region of interest (“ROI”) from a user, the system retrieves weather mapping data for the ROI and retrieves information related to attributes of structures within the ROI from a machine learning subsystem. The system then cross-references the property data, the weather data, and the structure attributes and assigns a risk rating to the structures within the ROI. Finally, the system generates and delivers a data package to the user.

Abstract: Systems and methods for controlling automated systems, such as robots and other devices/resources, using integer programming and column generation techniques, are provided. The system retrieves location and movement data corresponding to at least one automated system, and processes the location and movement data using an integer linear programming algorithm. The system also optimizes the integer linear programming algorithm using a column generation algorithm, and solves a pricing problem associated with generating routes for the at least one automated system. Then, the system generates a route based on outputs of the integer linear programming algorithm, the column generation algorithm, and solving of the pricing problem, and transmits the route to the at least one automated system, the at least one automated system executing the route to control a location or a movement of the at least one automated system.

Abstract: Machine learning systems and methods for multiscale Alzheimer's dementia recognition through spontaneous speech are provided. The system retrieves one or more audio samples and processes the one or more audio samples to extract acoustic features from audio samples. The system further processes the one or more audio samples to extract linguistic features from the audio samples. Machine learning is performed on the extracted acoustic and linguistic features, and the system indicates a likelihood of Alzheimer's disease based on output of machine learning performed on the extracted acoustic and linguistic features.

Abstract: Systems and methods for rapidly developing annotated computer models of structures and properties is provided. The system generates three-dimensional (3D) models of structures and property using a wide variety of digital imagery, and/or can process existing 3D models created by other systems. The system processes the 3D models to automatically identify candidate objects within the 3D models that may be suitable for annotation, such as roof faces, chimneys, windows, gutters, etc., using computer vision techniques to automatically identify such objects. Once the candidate objects have been identified, the system automatically generates user interface screens which gather relevant information related to the candidate objects, so as to rapidly obtain, associate, and store annotation information related to the candidate objects.