Abstract: The present approach relates to the use of machine learning and deep learning systems suitable for solving large-scale, space-variant tomographic reconstruction and/or correction problems. In certain embodiments, a tomographic transform of measured data obtained from a tomography scanner is used as an input to a neural network. In accordance with certain aspects of the present approach, the tomographic transform operation(s) is performed separate from or outside the neural network such that the result of the tomographic transform operation is instead provided as an input to the neural network. In addition, in certain embodiments, one or more layers of the neural network may be provided as wavelet filter banks.
Type:
Grant
Filed:
April 5, 2017
Date of Patent:
November 12, 2019
Assignee:
GENERAL ELECTRIC COMPANY
Inventors:
Lin Fu, Sathish Ramani, Bruno Kristiaan Bernard De Man, Xue Rui, Sangtae Ahn
Abstract: A method includes a computing device a computer device receiving extracted features from a driver-facing camera and from a road as viewed by a road-facing camera; the computer device further receiving extracted features reflecting the driver's behavior including head and eyes movement, speech and gestures; the computer device further receiving extracted telemetry features from a vehicle; the computer device still further receiving extracted features reflecting the driver's biometrics; and a decision engine receiving information from the computer device representing each of the extracted features of the driver, wherein a driver's attention and emotional state is determined to evaluate risks associated to moving vehicles and the driver's ability to deal with any projected risks.
Abstract: The present disclosure relates to a phenotype analysis of cellular image data using a deep metric network. One example embodiment includes a method. The method includes receiving, by a computing device, a plurality of candidate images of candidate biological cells each having a respective candidate phenotype. The method also includes obtaining, by the computing device for each of the plurality of candidate images, a semantic embedding associated with the respective candidate image. Further, the method includes grouping, by the computing device, the plurality of candidate images into a plurality of phenotypic strata based on their respective semantic embeddings.
Abstract: In an evaluation information provision system, subject's motion data is stored in association with attributes. When an attribute is assigned, the evaluation information provision system selects feature data from a plurality of the subject's motion data associated with the assigned attribute. The evaluation information provision system calculates a score for the assigned attribute, for the user's motion, using a statistical distance between the selected feature data and the user's motion data. The calculated score is displayed, for example, as a screen.
Type:
Grant
Filed:
October 18, 2016
Date of Patent:
October 29, 2019
Assignees:
UNIVERSITY OF TSUKUBA, OSAKA UNIVERSITY, MIZUNO CORPORATION
Abstract: A wireless sound-emitting device includes a housing adapted to be coupled to a wall at a source of electric power, a loudspeaker positioned at a periphery of the housing, a control module outputting an electric audio signal to the at least one loudspeaker, and a wireless communications module in electrical communication with the control module. The loudspeaker emits acoustic signals in a direction parallel to the wall, when the housing is coupled to the wall, with the acoustic signals reflecting off the wall. The device may produce a sound masking noise or play a sound recorded on an internal memory. The device may include an electric plug or be adapted to replace an electric outlet faceplate. The device may have electric pass-through outlets and may be powered by the source of electric power. The device may be controlled remotely, for example via an Internet of Things (IoT) platform.
Type:
Grant
Filed:
March 22, 2018
Date of Patent:
October 22, 2019
Assignee:
Cambridge Sound Management, Inc.
Inventors:
Christopher Calisi, Gordon V. Cook, Edward Driscoll, Thomas R. Horrall
Abstract: Provided are a method and system for identifying and classifying the owner, age and health of an optic nerve head and its vasculature based on analysis of vector relationships of blood vessels and the neuroretinal rim within an image of the optic nerve head to each other.
Abstract: A system and methods related to hitch assist are provided herein. An imager captures rear-vehicle images. A processor is configured to extract trailer and ground features from the captured images, compute vehicle motion displacement based on optical flow of the ground features, estimate a height of each trailer feature based on vehicle motion displacement and optical flow of the trailer features, and determine a trailer height based on the estimated heights of at least a portion of the trailer features.
Type:
Grant
Filed:
February 22, 2018
Date of Patent:
October 1, 2019
Assignee:
Ford Global Technologies, LLC
Inventors:
Yi Zhang, Dongran Liu, Vidya Nariyambut murali
Abstract: Systems and methods of the present disclosure can facilitate determining a three-dimensional surface representation of an object. In some embodiments, the system includes a computer, a calibration module, which is configured to determine a camera geometry of a set of cameras, and an imaging module, which is configured to capture spatial images using the cameras. The computer is configured to determine epipolar lines in the spatial images, transform the spatial images with a collineation transformation, determine second derivative spatial images with a second derivative filter, construct epipolar plane edge images based on zero crossings of second derivative epipolar planes image based on the epipolar lines, select edges and compute depth estimates, sequence the edges based on contours in a spatial edge image, filter the depth estimates, and create a three-dimensional surface representation based on the filtered depth estimates and the original spatial images.
Abstract: The subject matter of the present disclosure generally relates to techniques for image analysis. In certain embodiments, various morphological or intensity-based features as well as different thresholding approaches may be used to segment the subpopulation of interest and classify object in the images.
Type:
Grant
Filed:
September 13, 2017
Date of Patent:
September 24, 2019
Assignee:
GENERAL ELECTRIC COMPANY
Inventors:
Alberto Santamaria-Pang, Qing Li, Yunxia Sui, Dmitry Vladimirovich Dylov, Christopher James Sevinsky, Michael E. Marino, Michael J. Gerdes, Daniel Eugene Meyer, Fiona Ginty, Anup Sood
Abstract: Techniques described herein may involve dynamically adjusting the equalization of multiple speakers in a playback device. An example implementation involves a playback device emitting, via a speaker, a first audio signal. While emitting the first audio signal, the playback device detects, via a microphone, data representing a second audio signal, the second audio signal comprising one or more reflections of the first audio signal as emitted by the speaker within a surrounding environment. The playback device determines a frequency response curve based on the second audio signal, the frequency response curve representing acoustic characteristics of a system comprising the playback device and the surrounding environment. The playback device determines filter coefficients based on a difference between the determined frequency response curve and a target frequency response curve, and applies the filter coefficients to one or more filters during playback of audio content by the playback device.
Abstract: The present invention provides methods and systems for digitally processing audio signals in two-channel audio systems and/or applications. In particular, the present invention includes a first filter structured to split a two-channel audio input signal into a low frequency signal and a higher frequency signal. An M/S splitter is then structured to split the higher frequency signal into a middle and a side signal. A detection module is then configured to create a detection signal from the middle signal, which is used in a compression module configured to modulate the side signal to create a gain-modulated side signal. A processing module is then structured to combine the low frequency signal, middle signal, and the gain-modulated side signal to form a final output signal.
Type:
Grant
Filed:
January 30, 2018
Date of Patent:
September 10, 2019
Assignee:
Bongiovi Acoustics LLC
Inventors:
Anthony Bongiovi, Glenn Zelniker, Joseph G. Butera, III
Abstract: A system and method for managing wireless earpieces. Circuitry of the wireless earpieces are powered utilizing a high-power mode in response to detecting a magnetic field is not applied to one or more of the wireless earpieces. The power sent to the circuitry of the wireless earpieces is altered to a low power mode in response to detecting the magnetic field is applied to one or more of the wireless earpieces.
Type:
Grant
Filed:
September 27, 2018
Date of Patent:
August 27, 2019
Assignee:
BRAGI GmbH
Inventors:
Eric Christian Hirsch, Peter Vincent Boesen
Abstract: A fingerprint identification system, comprising panel, for placing finger; light source, disposed under panel, for generating incident light, wherein incident light is emitted to and reflected by finger to generate reflected light; processor, for performing fingerprint identification on finger according to plurality of sensing signals; and optical sensing array, disposed under panel and coupled to processor, comprising plurality of sensing units for generating plurality of sensing signals, each of sensing units comprising optical sensor, for receiving reflected light reflected by finger to generate sensing signal; focusing layer, disposed under panel, for focusing reflected light reflected by finger; blocking layer, disposed under focusing layer, for blocking part of reflected light to reduce diffraction phenomenon of reflected light; shading layer, disposed under blocking layer, for blocking part of reflected light to limit incident angle of reflected light; and aperture layer, disposed under shading layer a
Type:
Grant
Filed:
January 15, 2018
Date of Patent:
August 20, 2019
Assignee:
Egis Technology Inc.
Inventors:
Yu-Hsiang Huang, Chen-Chih Fan, Tong-Long Fu
Abstract: A method for vehicle control includes receiving a color image from an imaging system. The color image includes a traffic indicator with a color portion. The method includes extracting red color components from the color image by subtracting a grayscale intensity value of each pixel from a red-scale value of each pixel. The method includes extracting green color components from the color image by subtracting the grayscale intensity value of each pixel from a green-scale value of each pixel. The method includes performing blob analysis based on the red color components and the green color components. The method includes determining, based on the blob analysis, a color of the color portion of the traffic indicator, and controlling a vehicle system of a vehicle based on the color of the color portion of the traffic indicator.
Type:
Grant
Filed:
March 6, 2017
Date of Patent:
August 13, 2019
Assignee:
Honda Motor Co., Ltd.
Inventors:
Kin C. Fung, Timothy J. Dick, Dhanashree Palande
Abstract: The image processing apparatus includes a processing circuitry. The processing circuitry is configured to obtain first volume data showing a morphological shape of an object and second volume data showing information corresponding to a position spatially different from a position of the object. The processing circuitry is configured to provide the information in the second volume data to the position of the object in the first volume data.
Abstract: A method of operating a package handling system includes: receiving first image data representative of a first surface image of a package surface; identifying a pair of intersecting edges for the package surface based on the first image; determining a minimum viable region based on the pair of edges, the minimum viable region for gripping and lifting the package; receiving second image data representative of the package after the lift; and creating registration data based on the third image data.
Type:
Grant
Filed:
March 1, 2019
Date of Patent:
August 6, 2019
Assignee:
MUJIN, Inc.
Inventors:
Rosen Diankov, Huan Liu, Xutao Ye, Jose Jeronimo Moreira Rodrigues, Yoshiki Kanemoto, Jinze Yu, Russell Islam
Abstract: Methods and systems are provided for generating quantitative computed tomography (CT) angiographic images using imaging systems that acquire a set of projection views forming CT angiographic image data of a patient using a projection duration of less than 50 milliseconds. The method may include producing a composite image from the CT angiographic image data that indicates an attenuation value at each composite image pixel of the patient, backprojecting each projection view in the CT angiographic image data and weighting a value backprojected into at image pixel by an attenuation value of a corresponding pixel in the composite image and summing backprojected values for each image pixel to produce a CT image of the patient. The method may further include performing a scatter correction and determining at least one of a flow direction or a velocity of flow with a vessel in the patient to provide the quantitative CT angiographic images.
Abstract: Example systems disclosed herein to detect watermark modifications include a watermark encoder to encode a second watermark in a sampled media signal obtained from a received broadcast signal, the sampled media signal already encoded with a first watermark that was included in the received broadcast signal. Disclosed example systems also include a watermark decoder to detect the first watermark and the second watermark in the sampled media signal, and a watermark modification evaluator to compare a first metric determined for the first watermark and a second metric determined for the second watermark to determine whether the first watermark was modified prior to being included in the received broadcast signal. Disclosed example systems further include a ratings server to revise ratings data corresponding to the received broadcast signal when the first watermark is determined to have been modified prior to being included in the received broadcast signal.
Abstract: A method (100) for comparing a first video shot (Vs1) comprising a first set of first images (I1(s)) with a second video shot (Vs2) comprising a second set of second images (I2(t)), at least one between the first and the second set comprising at least two images.
Abstract: A system according to various exemplary embodiments includes a processor and a user interface coupled to the processor, the user interface comprising an input device and a display screen. The system further comprises a sensor component coupled to the processor, the sensor component comprising a location sensor for determining location information associated with the system, and memory coupled to the processor and storing instructions that, when executed by the processor, cause the system to perform operations comprising: retrieving sensor information from the sensor component, the sensor information including location information from the location sensor; retrieving, from the memory, avatar information associated with a user of the system; generating a customized graphic based on the sensor information and the user's avatar information; and presenting the customized graphic within an electronic message via the display screen of the user interface.
Type:
Grant
Filed:
October 11, 2016
Date of Patent:
July 9, 2019
Assignee:
Snap Inc.
Inventors:
Dorian Franklin Baldwin, Jacob Blackstock, Shahan Panth, David James Kennedy