SPLITTING OF INPUT DATA FOR PROCESSING IN NEURAL NETWORK PROCESSOR

    公开(公告)号:US20250165784A1

    公开(公告)日:2025-05-22

    申请号:US19030424

    申请日:2025-01-17

    Applicant: APPLE INC.

    Abstract: Embodiments of the present disclosure relate to splitting input data into smaller units for loading into a data buffer and neural engines in a neural processor circuit for performing neural network operations. The input data of a large size is split into slices and each slice is again split into tiles. The tile is uploaded from an external source to a data buffer inside the neural processor circuit but outside the neural engines. Each tile is again split into work units sized for storing in an input buffer circuit inside each neural engine. The input data stored in the data buffer and the input buffer circuit is reused by the neural engines to reduce re-fetching of input data. Operations of splitting the input data are performed at various components of the neural processor circuit under the management of rasterizers provided in these components.

    USER INTERFACES FOR DEVICES WITH MULTIPLE DISPLAYS

    公开(公告)号:US20250165207A1

    公开(公告)日:2025-05-22

    申请号:US19030689

    申请日:2025-01-17

    Applicant: Apple Inc.

    Abstract: In some embodiments, an electronic device displays and allows user interaction with content on multiple displays. In some embodiments, an electronic device displays three-dimensional content using coordinated views on multiple displays. In some embodiments, an electronic device shares content while displaying a video conferencing user interface. In some embodiments, an electronic device facilitates the display or keeping private of content on multiple displays. In some embodiments, an electronic device facilitates the display of information from applications using a shared graphical input coordinate space. In some embodiments, an electronic device presents indications of notifications. In some embodiments, an electronic device presents representations of items of content that are related to content presented on the one or more electronic devices. In some embodiments, an electronic device presents user interfaces including a soft keyboard.

    Multi-Finger Gesture based on Finger Manipulation Data and Extremity Tracking Data

    公开(公告)号:US20250165070A1

    公开(公告)日:2025-05-22

    申请号:US19027964

    申请日:2025-01-17

    Applicant: Apple Inc.

    Abstract: A method is performed at an electronic device with one or more processors, a non-transitory memory, a display, an extremity tracking system, and a communication interface provided to communicate with a finger-wearable device. The method includes displaying a computer-generated object on the display. The method includes obtaining finger manipulation data from the finger-wearable device via the communication interface. The method includes determining a multi-finger gesture based on extremity tracking data from the extremity tracking system and the finger manipulation data. The method includes registering an engagement event with respect to the computer-generated object according to the multi-finger gesture.

    Grouping Maneuvers For Display In A Navigation Presentation

    公开(公告)号:US20250164269A1

    公开(公告)日:2025-05-22

    申请号:US19029038

    申请日:2025-01-17

    Applicant: Apple Inc.

    Abstract: Some embodiments of the invention provide several novel methods for generating a navigation presentation that displays a device navigating a route on a map. The method of some embodiments uses a virtual camera that, based on detected changes in the navigation context, dynamically modifies the way it captures portions of the map to produce different navigation scenes in the navigation presentation. To generate the navigation scenes, the method of some embodiments (1) identifies different sets of attributes that describe the different navigation contexts at different times during the navigation presentation, and (2) uses these different sets of attributes to identify different styles for operating the virtual camera. In some embodiments, the method uses an identified style to specify the virtual camera's positional attributes, which, in turn, define the portions of the map that the virtual camera identifies for rendering to produce several navigation scenes for a period of time (e.g., until the navigation context changes, or until the navigation presentation ends when the navigation context does not change again). During the navigation presentation, each time the navigation context changes, the identified set of attributes may change. This change, in turn, may cause the method of some embodiments to select a new style for operating the virtual camera. When the style for operating the virtual camera changes. the method of some embodiments modifies the way the virtual camera captures the portion of the map to render.

    ATTACHMENT SYSTEM FOR WATCHBAND
    19.
    发明申请

    公开(公告)号:US20250160489A1

    公开(公告)日:2025-05-22

    申请号:US19030698

    申请日:2025-01-17

    Applicant: Apple Inc.

    Abstract: Wearable electronic devices, such as watches, can be provided with an attachment system for connecting and releasing a removable module that interfaces with some portion of the electronic device. For example, a watch housing includes a main body having a channel for receiving a watchband and a first opening extending to the channel. A rear cover coupled to the main body provides a second opening. A unibody button is positioned between the main body and the rear cover. The button is biased against the rear cover and is moveable toward the channel to release a watchband engaged therein.

Patent Agency Ranking