Ford Global Technolgoies, LLC

University of Michigan

Hafiz Malik

Ali Hassani

Jonathan Diedrich

Ryan Edwin Hanson

Justin Miller, Jr

John Moore

Zaid El Shair

John Robert Van Wiemeersch

Robert Parenti

Alexandra Taylor

Francis Obiagwu

Amit Kulkarni

David Hiskens

Priyank Shah

Jason Carlton

Muteb Aljasem

Rafi Ud Daula Refat

Syed Irtaza

Methods and systems are provided for deleting a user profile and pairing data stored in memory of an in-vehicle computing system to maintain user privacy and data security. In one method, the method includes deleting at least one user profile and corresponding pairing data upon receiving a deletion request, wherein the in-vehicle computing system is configured to operate in a rental mode or a non-rental mode, and wherein a pre-set trigger varies according to the rental mode or the non-rental mode. In another method, the method includes deleting at least one user profile and corresponding pairing data based on one or more of a pre-determined location and a status of an ignition of the vehicle in a rental mode and deleting the at least one user profile and the corresponding pairing data based on a prompt displayed to a user in a non-rental mode.

Systems and methods for securing private user data based on GPS

Output can be provided from a selected biometric analysis task that is one of a plurality of biometric analysis tasks based on an image provided from an image sensor. The selected biometric analysis task can be performed in a deep neural network that includes a common feature extraction neural network, a plurality of biometric task-specific neural networks and a plurality of expert pooling neural networks that perform the plurality of biometric analysis tasks by inputting the image to the common feature extraction network to determine latent variables. The latent variables can be input to the plurality of biometric task-specific neural networks to determine a plurality of first outputs. Concatenated first output results can be formed and the concatenated plurality of first result outputs and the latent variables can be input to the plurality of expert pooling neural networks to determine one or more biometric analysis task outputs.

Biometric task network

Traditional speaker verification systems are vulnerable to voice spoofing attacks, such as voice-replay attack, voice-cloning attack, and cloned-replay attack. To overcome these vulnerabilities, a secure automatic speaker verification system based on a novel sign modified acoustic local ternary pattern (sm-ALTP) features and asymmetric bagging-based classifier-ensemble with enhanced attack vector is presented. The proposed audio representation approach clusters the high and low frequency components in audio frames by normally distributing them against a convex function. Afterwards, the neighborhood statistics are applied to capture the user specific vocal tract information.

Secure automatic speaker verification system

A facial recognition authentication system and method include obtaining a current image in response to a request for facial recognition authentication of a user. A set of current image features are generated from the input image. Live image features and spoof image features are retrieved for the user. Then a determination is output that the input image is one of live or spoofed based on a comparison of the current image features to the live image features and the spoof image features.

Spoof images for user authentication

A deep neural network can provide output from a selected biometric analysis task that is one of a plurality of biometric analysis tasks based on an image. The selected biometric analysis task can be performed in a deep neural network that includes a common feature extraction neural network, a plurality of biometric task-specific neural networks, a plurality of segmentation mask neural networks and an expert pooling neural network that perform the plurality of biometric analysis tasks by inputting the image to the common feature extraction network to determine latent variables. The latent variables can be input to the plurality of biometric task-specific neural networks to determine a plurality of biometric analysis task outputs. The latent variables can be input to a segmentation neural network to determine a facial feature segmentation output. The facial feature segmentation output can be output to a plurality of segmentation mask neural networks.

A computer, including a processor and a memory, the memory including instructions to be executed by the processor to acquire one or more images from a camera and determine first camera noise values based on the one or more images by determining reactions of camera photo receptors to light. The instructions can include further instructions to compare the first camera noise values with second camera noise values determined based on previously acquired images from the camera and output a tamper determination for the camera based on whether the first camera noise values match, within a tolerance value, the second camera noise values determined based on the previously acquired images from the camera.

Camera tampering detection

A computer, including a processor and a memory, the memory including instructions to be executed by the processor to acquire a first image with a visible and NIR light camera and acquire a second image with an infrared camera. The instructions can include further instructions to determine whether the second image includes a live human face by comparing a first infrared profile included in the second image with second infrared profile included in a previously acquired third image acquired with the infrared camera; and when the second image includes the live human face, output the first image.

Counterfeit image detection

A computer, including a processor and a memory, the memory including instructions to be executed by the processor to acquire a first image from a first camera by illuminating a first object with a first light and determine an object status as one of a real object or a counterfeit object by comparing a first measure of pixel values corresponding to the first object to a threshold.

A computer, including a processor and a memory, the memory including instructions to be executed by the processor to divide each of one or more images acquired by a camera into a plurality of zones, determine respective camera noise values for respective zones based on the one or more images, determine one or more zone expected values for the one or more images by summing camera noise values multiplied by scalar coefficients for each zone and normalizing the sum by dividing by a number of zones in the plurality of zones, and determine a source of the camera as being one of the same camera or an unknown camera based on comparing the one or more zone expected values to previously acquired expected zone values.

Camera identification

A computer includes a processor and a memory, the memory storing instructions executable by the processor to collect a digital image that includes a plurality of pixels with a first sensor, input a reference data string, a key data string, and a set of collected data from a second sensor into a permutation generator that outputs a watermark data string, and embed the watermark data string in the digital image at specified pixels in the plurality of pixels.

Enhanced vehicle operation

A vehicle computer includes a watermark memory and a watermark processor programmed to execute instructions stored in the watermark memory. The instructions executed by the watermark processor include receiving an image captured by a camera, selecting a set of random pixel locations, generating a random watermark, and embedding the random watermark into the image at the set of random pixel locations. Another vehicle computer includes a validation memory and a validation processor programmed to execute instructions stored in the validation memory. The instructions executed by the validation processor include receiving a watermarked image, determining a random watermark, detecting an embedded watermark in the received watermarked image by selecting a set of random pixels and analyzing the selected set of random pixels for the random watermark, and authenticating the watermarked image as a result of determining that the watermarked image includes the random watermark at the set of random pixel locations.

Growing community of inventors

Canton, MI, United States of America

Hafiz Malik