Apple Inc.

Thorsten Gernoth

Feng Tang

Haitao Guo

Kelsey Yukio Ho

Onur Canturk Hamsici

Ian R Fasel

Atulit Kumar

Steven Porter Hotelling

Jerrold Von Hauck

Moshe Laifenfeld

Anup K Sharma

Xiaojin Shi

Lucia Elena Ballard

Marcel Van Os

Steve Porter Hotelling

Lynn Richard Youngs

Ivan Krstic

Jonathan Pokrass

Thomas P Mensch

Michael Ross Malone

Matthew C Waldon

Tal Kaitz

Petr Kostka

Andrei Kolin

Deepti S Prakash

Christopher G Zeleznik

Chong Chen

Gideon Littwin

Eitan Hirsh

Ziv Hendel

Etai Littwin

Lucie Kucerova

Srikrishna Sridhar

Shawn R Scully

Touraj Tajbakhsh

Carlos E Guestrin

Eslam A Mostafa

Shreyas V Joshi

Pavan Kumar Anasosalu Vasu

Gustav M Larsson

Leon A Gatys

Jia-Chi Wu

Kory R Watson

Karla P Vega

Shuo Feng

Michael P Johnson

Allison W Dryer

Zachary Z Becker

Michelle Chua

Hao Tang

Joerg A Liebelt

Onur C Hamsici

Pavan Kumar Ansosalu Vasu

Cheng Lu

Donghoon Lee

Deepti Sunder Prakash

Chris Zeleznik

An exemplary process acquires frames of image data in a physical environment including an object, tracks an orientation of the device with respect to the object during the acquiring of the frames, selects keyframes from the frames based on detecting a first condition and selecting a first frame as a keyframe based on a first keyframe selection criterion and the orientation of the device at the first frame, and based on detecting a second condition, selecting a second frame as a keyframe based on a second keyframe selection criterion and the orientation of the device at the second frame, wherein the second keyframe selection criterion is different than the first keyframe selection criterion, and generates an object scan (e.g., a pose graph) based on the frames of image data and the selected keyframes.

Dynamic keyframing in object centric pose estimation

Various implementations disclosed herein provide feedback to a user during object scanning based on how well sensor data of the scanned object has been captured. During object scanning a user may move an electronic device with sensors (e.g., cameras, depth sensors, etc.) around an object to capture sensor data for use in generating a final 3D model of the object. Live feedback during the scanning process is enabled by assessing how well the captured sensor data represents different portions of the object. In some implementations, a 3D model is generated, updated, and assessed based on the sensor data live during the scanning process. This live 3D model may be coarser (i.e., having fewer details) than the final 3D model.

Feedback using coverage for object scanning

Generating a three-dimensional virtual representation of a three-dimensional physical object can be based on capturing or receiving a capture bundle or a set of images. In some examples, generating the virtual representation of the physical object can be facilitated by user interfaces for identifying a physical object and capturing a set of images of the physical object. Generating the virtual representation can include previewing or modifying a set of images. In some examples, generating the virtual representation of the physical object can include generating a first representation of the physical object (e.g., a point cloud) and/or generating a second three-dimensional virtual representation of the physical object (e.g., a mesh reconstruction). In some examples, a visual indication of the progress of the image capture process and/or the generation of the virtual representation of the three-dimensional object can be displayed, such as in a capture user interface.

Systems, methods, and user interfaces for generating a three-dimensional virtual representation of an object

Devices, systems, and methods are disclosed for partial point cloud registration. In some implementations, a method includes obtaining a first set of three-dimensional (3D) points corresponding to an object in a physical environment, the first set of 3D points having locations in a first coordinate system, obtaining a second set of 3D points corresponding to the object in the physical environment, the second set of 3D points having locations in a second coordinate system, predicting, via a machine learning model, locations of the first set of 3D points in the second coordinate system, and determining transform parameters relating the first set of 3D points and the second set of 3D points based on the predicted location of the first set of 3D points in the second coordinate system.

Deep partial point cloud registration of objects

A device implementing a system for providing predicted RGB images includes at least one processor configured to obtain an infrared image of a subject, and to obtain a reference RGB image of the subject. The at least one processor is further configured to provide the infrared image and the reference RGB image to a machine learning model, the machine learning model having been trained to output predicted RGB images of subjects based on infrared images and reference RGB images of the subjects. The at least one processor is further configured to provide a predicted RGB image of the subject based on output by the machine learning model.

Machine learning assisted image prediction

Techniques are disclosed relating to biometric authentication, e.g., facial recognition. In some embodiments, a device is configured to verify that image data from a camera unit exhibits a pseudo-random sequence of image capture modes and/or a probing pattern of illumination points (e.g., from lasers in a depth capture mode) before authenticating a user based on recognizing a face in the image data. In some embodiments, a secure circuit may control verification of the sequence and/or the probing pattern. In some embodiments, the secure circuit may verify frame numbers, signatures, and/or nonce values for captured image information. In some embodiments, a device may implement one or more lockout procedures in response to biometric authentication failures. The disclosed techniques may reduce or eliminate the effectiveness of spoofing and/or replay attacks, in some embodiments.

Biometric authentication techniques

An operation of a facial recognition authentication process may fail to authenticate a user even if the user is an authorized user of the device. In such cases, the facial recognition authentication process may automatically re-initiate to provide another attempt to authenticate the user using additional captured images. For the new attempt (e.g., the retry) to authenticate the user, one or more criteria for the images used in the facial recognition authentication process may be adjusted. For example, criteria for distance between the camera and the user's face and/or occlusion of the user's face in the images may be adjusted before the new attempt to authenticate the user. Adjustment of these criteria may increase the likelihood that the authorized user will be successfully authenticated in the new attempt.

Automatic retries for facial recognition

Intelligent systems are disclosed that respond to user intent and desires based upon activity that may or may not be expressly directed at the intelligent system. In some embodiments, the intelligent system acquires a depth image of a scene surrounding the system. A scene geometry may be extracted from the depth image and elements of the scene may be monitored. In certain embodiments, user activity in the scene is monitored and analyzed to infer user desires or intent with respect to the system. The interpretation of the user's intent as well as the system's response may be affected by the scene geometry surrounding the user and/or the system. In some embodiments, techniques and systems are disclosed for interpreting express user communication, e.g., expressed through hand gesture movements. In some embodiments, such gesture movements may be interpreted based on real-time depth information obtained from, e.g., optical or non-optical type depth sensors.

Multi media computing or entertainment system for responding to user presence and activity

Occlusion of facial features may be detected and assessed in an image captured by a camera on a device. Landmark heat maps may be used to estimate the location of landmarks such as the eyes, mouth, and nose of a user's face in the captured image. An occlusion heat map may also be generated for the captured image. The occlusion heat map may include values representing the amount of occlusion in regions of the face. The estimated locations of the eyes, mouth, and nose may be used in combination with the occlusion heat map to assess occlusion scores for the landmarks. The occlusion scores for the landmarks may be used control one or more operations of the device.

Occlusion detection for facial recognition processes

A neural network may implement a face detection process on a mobile device or computer system. An image captured using a camera on the device may be processed using the face detection process. The face detection process may provide a bounding box for the face detected in the image. The face detected in the image may have any orientation. The face detected in the image may include either a whole face of the user or only a partial face of the user. The face detection process may provide face detection that is substantially invariant to exposure in the captured image. The face may also be detected over a wide range of distances between the camera and the face of the user.

Robust face detection

Detection of a user paying attention to a device may be used to enable or support biometric security (e.g., facial recognition) enabled features on the device. Images captured by a camera on the device may be used to determine if the user is paying attention to the device. Facial features of the user's face in the images may be assessed to determine if the user is paying attention to the device. Facial features may be assessed through comparison of feature vectors generated from the captured image feature to a set of known feature vectors. The known feature vectors for attention may be generated using a machine learning process.

Attention detection

Growing community of inventors

San Francisco, CA, United States of America

Thorsten Gernoth