State Farm Mutual Automobile Insurance Company

Michael L Bernico

Jeffrey S Myers

Kenneth Jason Sanchez

Marigona Bokshi-Drotar

Tim G Sanidas

Edward W Breitweiser

Dingchao Zhang

Utku Pamuksuz

Peter Laube

Bradley A Sliz

Holly Lambert

Michael K Cook

He Yang

Shane Tomlinson

Jennifer Malia Andrus

Daniel J Green

V Rao Kanneganti

Lynn Calvo

Alex Anton

Jeff Perschall

Milton Mowery

Christopher Zimmer

Matthew E Vose

Brian Mark Fields

Scott T Christensen

Nathan Lee Tofte

Steven C Cielocha

Kristopher Keith Gaudin

Steve Roberson

John Adrian Nepomuceno

Duane Christiansen

Abigail A Scott

Brian Alexander

Andrew J Rader

A method of training and using a machine learning model that controls for consideration of undesired factors which might otherwise be considered by the trained model during its subsequent analysis of new data. For example, the model may be a neural network trained on a set of training images to evaluate an insurance applicant based upon an image or audio data of the insurance applicant as part of an underwriting process to determine an appropriate life or health insurance premium. The model is trained to probabilistically correlate an aspect of the applicant's appearance with a personal and/or health-related characteristic. Any undesired factors, such as age, sex, ethnicity, and/or race, are identified for exclusion. The trained model receives the image (e.g., a 'selfie') of the insurance applicant, analyzes the image without considering the identified undesired factors, and suggests the appropriate insurance premium based only on the remaining desired factors.

Method of controlling for undesired factors in machine learning models

A method and system may use computer vision techniques and machine learning analysis to automatically identify a user's biometric characteristics. A user's client computing device may capture a video of the user. Feature data and movement data may be extracted from the video and applied to statistical models for determining several biometric characteristics. The determined biometric characteristic values may be used to identify individual health scores and the individual health scores may be combined to generate an overall health score and longevity metric. An indication of the user's biometric characteristics which may include the overall health score and longevity metric may be displayed on the user's client computing device.

Biometric characteristic application using audio/video analysis

A method of training and using a machine learning model that controls for consideration of undesired factors which might otherwise be considered by the trained model during its subsequent analyses of new data. For example, the model may be a neural network trained on a set of training images to evaluate an insurance applicant based upon an image or audio data of the insurance applicant as part of an underwriting process to determine an appropriate life or health insurance premium. The model is trained to probabilistically correlate an aspect of the applicant's appearance with a personal and/or health-related characteristic. Any undesired factors, such as age, sex, ethnicity, and/or race, are identified for exclusion. The trained model receives the image (e.g., a 'selfie') of the insurance applicant, analyzes the image without considering the identified undesired factors, and suggests the appropriate insurance premium based only on the remaining desired factors.

A system and computer-implemented method for facilitating a user of a mobile device obtaining image data of damage to a vehicle for damage assessment includes capturing image data of a vehicle with the mobile device. The mobile device may include an orientation model for capturing the image data. The captured image data is analyzed, and a determination is made of the orientations of the images of the captured image data. In addition, a determination is made as to whether the captured image data can be used for the damage assessment. The captured image data may then be transmitted to a damage estimator computing device for estimating an amount of damage to the vehicle.

Methods and systems for obtaining image data of a vehicle for automatic damage assessment

A system and computer-implemented method for processing images of a damaged vehicle and estimating a repair cost of the damaged vehicle includes receiving image data of a vehicle from a user mobile device. The image data is processed to determine whether one or more parts of the damaged vehicle are damaged. In addition, one or more parts of the damaged vehicle are identified for repair and one or more parts of the damaged vehicle are identified for replacement. A cost associated for the repair of each of the one or more parts of the damaged vehicle for repair is estimated based on estimated repair cost data contained in a parts repair database. Moreover, a cost associated with the replacement of each of the one or more parts of the damaged vehicle for replacement is determined based on replacement cost data contained in a parts replacement database.

Methods and systems for automatic processing of images of a damaged vehicle and estimating a repair cost

A system and method for evaluating an insurance applicant as part of an underwriting process to determine one or more appropriate terms of life or other insurance coverage, such as premiums. A processing element employing a neural network is trained to correlate aspects of appearance and/or voice with personal and/or health-related characteristic. A database of images and/or voice recordings of individuals with known personal and/or health-related characteristics is provided for this purpose. The processing element is then provided with an image and/or voice recording of the insurance applicant. The image may be an otherwise non-diagnostic image, such as an ordinary 'selfie.' The trained processing element analyzes the image of the insurance applicant, with their permission or affirmative consent, to determine the personal and/or health-related characteristic for the insurance applicant, and then, based upon that analysis, facilitates the underwriting process and/or suggests the one or more appropriate terms of insurance coverage.

Using images and voice recordings to facilitate underwriting life insurance

A method and system may use machine learning analysis of audio data to automatically identify a user's biometric characteristics. A user's client computing device may capture audio of the user. Feature data may be extracted from the audio and applied to statistical models for determining several biometric characteristics. The determined biometric characteristic values may be used to identify individual health scores and the individual health scores may be combined to generate an overall health score and longevity metric. An indication of the user's biometric characteristics which may include the overall health score and longevity metric may be displayed on the user's client computing device.

Method and system for identifying biometric characteristics using machine learning techniques

Systems and methods for real-time detection and mitigation anomalous behavior of a remote vehicle are provided, e.g., vehicle behavior that is consistent with distracted or unexpectedly disabled driving. On-board and off-board sensors associated with a subject vehicle may monitor the subject vehicle's environment, and behavior characteristics of a remote vehicle operating within the subject vehicle's environment may be determined based upon collected sensor data. The remote vehicle's behavior characteristics may be utilized to detect or determine the presence of anomalous behavior, which may be anomalous for the current contextual conditions of the vehicles' environment. Mitigating actions for detected remote vehicle anomalous behaviors may be suggested and/or automatically implemented at the subject vehicle and/or at proximate vehicles to avoid or reduce the risk of accidents, injury, or death resulting from the anomalous behavior. In some situations, authorities may be notified.

Technology for real-time detection and mitigation of remote vehicle anomalous behavior

Systems and methods for identifying anomalous driving behavior for a vehicle based on past driving behavior are disclosed herein. The method may include receiving a set of time-series driving data for the vehicle, wherein the set of time-series driving data is indicative of a set of operating conditions for the vehicle. Performing machine learning operations on the set of time-series driving data. Identifying a set of anomalous conditions in the time-series driving data based on a result set produced by the machine learning operations, wherein the set of anomalous conditions are indicative of an anomalous vehicle behavior. Comparing the set of anomalous conditions to a set of historical time-series driving data for the vehicle. Generating a vehicle feedback based on the time-series driving data and the comparison of the set of anomalous conditions to the set of historical time-series driving data.

Detecting and mitigating local individual driver anomalous behavior

Growing community of inventors

Bloomington, IL, United States of America

Michael L Bernico