-
公开(公告)号：US20220405500A1
公开(公告)日：2022-12-22
申请号：US17304419
申请日：2021-06-21
Applicant: Google LLC
Inventor： Mayank Bhargava , Idris Syed Aleem , Yinda Zhang , Sushant Umesh Kulkarni , Rees Anwyl Simmons , Ahmed Gawish
IPC: G06K9/00 , G06T7/73 , G06K9/32 , G06K9/62 , G06T17/00 , G06T7/50 , G06T19/20 , G06T7/246 , G02C7/02
Abstract: A computer-implemented method includes receiving a two-dimensional (2-D) side view face image of a person, identifying a bounded portion or area of the 2-D side view face image of the person as an ear region-of-interest (ROI) area showing at least a portion of an ear of the person, and processing the identified ear ROI area of the 2-D side view face image, pixel-by-pixel, through a trained fully convolutional neural network model (FCNN model) to predict a 2-D ear saddle point (ESP) location for the ear shown in the ear ROI area. The FCNN model has an image segmentation architecture.
-
公开(公告)号：US11475592B2
公开(公告)日：2022-10-18
申请号：US16787298
申请日：2020-02-11
Applicant: GOOGLE LLC
Inventor： Idris Aleem , Mayank Bhargava , Rees Simmons
Abstract: An ear saddle point of a subject's ear is determined to produce specifications to fit a wearable apparatus to a subject's head. A boundary of the subject's ear in a profile image is determined using a first model. A map of the subject's ear is generated indicating probable locations of a two-dimensional ear saddle point. A most probable location of the two-dimensional ear saddle point is determined based on the probability map. The two-dimensional ear saddle point is projected onto a three-dimensional mesh surface representing the subject's head. A maximum depth of the three-dimensional mesh surface is determined in a defined region around the projected two-dimensional ear saddle point. The ear saddle point is computed based on the projected two-dimensional ear saddle point and the determined maximum depth. Specifications of a wearable apparatus are generated to fit the apparatus to the subject's head based on the computed ear saddle point.
-
公开(公告)号：US11157077B2
公开(公告)日：2021-10-26
申请号：US16376604
申请日：2019-04-05
Applicant: GOOGLE LLC
Inventor： Idris S. Aleem , Andrew S. Logan , Mayank Bhargava
Abstract: A method of tracking an eye of a user on a wearable heads-up display (WHUD) worn on the head of the user includes generating infrared light over an eye tracking period, scanning the infrared light over the eye, and detecting reflections of the infrared light from the eye. A motion parameter that is sensitive to motion of the WHUD is measured. Eye tracking is performed in a first mode that is based on glint for values of the motion parameter that fall within a first range of motion parameter values for which an error in measurement of glint position. Eye tracking is performed in a second mode that is based on glint-pupil vector for values of the motion parameter that fall within a second range of motion parameter values for which an error in measurement of glint position exceeds the error threshold. A head-mounted apparatus with eye tracking is disclosed.
-
14.发明授权公开(公告)号：US11093034B2
公开(公告)日：2021-08-17
申请号：US16668677
申请日：2019-10-30
Applicant: GOOGLE LLC
Inventor： Idris S. Aleem , Mayank Bhargava , Andrew S. Logan
Abstract: A method of tracking a gaze position of an eye in a target space in a field of view of the eye over an eye tracking period includes performing a plurality of scans of the eye with infrared light within the eye tracking period. Each scan includes generating infrared light signals over a scan period and projecting the infrared light signals from a plurality of virtual light projectors to the eye to form a plurality of illumination areas on the eye. Reflections of the infrared light signals from the eye are detected for each scan. The gaze position of the eye in the target space is determined from the detected reflections of the infrared light signals for each scan.
-
-
-
Search Results
14
records
(took 0.013 seconds)
