Abstract: A content management system streams media items to client devices. The content management system stores a master (or uploaded version) of a video file in a first format (e.g., a first resolution and compression format) and receives a request to stream the video file to a client device using a second format. The video segments from the master are transcoded and stored in a buffer/cache on the content management system and streamed therefrom to the client device. Just-in-time transcoding is performed by waiting until the amount of transcoded segments that are stored in the buffer and have not yet been streamed falls below a predefined threshold before fetching another chunk of video segment from block storage and transcoding them. Transcoding of the first part of the video file is enabled before the master is completely uploaded partitioning the upload units to place the upload boundaries between video segments.

Abstract: Systems and methods for machine learning based fraud detection in checks for positive pay applications are provided. Various embodiments of the present technology provide systems and methods for overcoming deficiencies in prior systems, particularly the lack of explainability and high false positive rates. Embodiments extract payee information using a trained model and compares it with issued data. This process generates discrete, explainable matching signals for identified mismatches. A decision learning feature stores human review decisions based on key check identifiers, automatically filtering fraud signals for recurring legitimate transactions. Additionally, image feature clustering techniques are employed to curate robust datasets representative of real-world production data for model training.

Abstract: A miniaturized active dual pixel stereo system and method for close range depth extraction includes a projector adapted to project a locally distinct projected pattern onto an image of a scene and a dual pixel sensor including a dual pixel sensor array that generates respective displaced images of the scene. A three-dimensional image is generated from the displaced images of the scene by projecting the locally distinct projected pattern onto the image of the scene, capturing the respective displaced images of the scene using the dual pixel sensor, generating disparity images from the respective displaced images of the scene, determining depth to each pixel of the disparity images, and generating the three-dimensional image from the determined depth to each pixel. A three-dimensional image of a user's hands generated by the active dual pixel stereo system may be processed by gesture recognition software to provide an input to an electronic eyewear device.

Abstract: A method for minimizing latency of moving objects in an augmented reality (AR) display device is described. In one aspect, the method includes determining an initial pose of a visual tracking device, identifying an initial location of an object in an image that is generated by an optical sensor of the visual tracking device, the image corresponding to the initial pose of the visual tracking device. rendering virtual content based on the initial pose and the initial location of the object, retrieving an updated pose of the visual tracking device, tracking an updated location of the object in an updated image that corresponds to the updated pose, and applying a time warp transformation to the rendered virtual content based on the updated pose and the updated location of the object to generate transformed virtual content.

Abstract: A method for reducing motion-to-photon latency for hand tracking is described. In one aspect, a method includes accessing a first frame from a camera of an Augmented Reality (AR) device, tracking a first image of a hand in the first frame, rendering virtual content based on the tracking of the first image of the hand in the first frame, accessing a second frame from the camera before the rendering of the virtual content is completed, the second frame immediately following the first frame, tracking, using the computer vision engine of the AR device, a second image of the hand in the second frame, generating an annotation based on tracking the second image of the hand in the second frame, forming an annotated virtual content based on the annotation and the virtual content, and displaying the annotated virtual content in a display of the AR device.

Abstract: Systems and methods for LLM assessment are disclosed herein. Embodiments may provide a quality assessment of an LLM that is based on the consistency of that LLM. Utilizing response sampling, pairwise similarity, and an uncertainty score, embodiments may measure response consistency, which can be correlated with output quality for an LLM.

Abstract: A method for configuring a digital light projector (DLP) of an augmented reality (AR) display device is described. A light source component of the DLP projector is configured to generate a single red-green-blue color sequence repetition per image frame. The AR display device identifies a color sequence of the light source component of the DLP projector and tracks a motion of the AR display device. The AR display device adjusts an operation of the DLP projector based on the single red-green-blue color sequence repetition, the color sequence of the light source component of the DLP projector, and the motion of the AR display device.

Abstract: A depth estimation system to perform operations that include: receiving image data generated by a client device, the image data comprising a depiction of an environment; identifying a set of image features based on the image data; determining a pose of the client device based on the set of features; generating a depth estimation based on the image data and the pose of the client device; and generating a mesh model of the environment based on the depth estimation.

Abstract: A battery pack supplies an electrically driven treatment apparatus with an electric driving power. The battery pack includes: a stack housing, wherein the stack housing has a common housing opening defined by a housing edge, and a plurality of pouch cells, wherein the pouch cells have cell tabs and are configured and disposed in a stack within the stack housing such that the cell tabs at least with tab parts on the common housing opening project beyond at least one edge portion of the housing edge. The tab parts are directly electrically connected to each other by welded connections.

Abstract: A method for reducing motion-to-photon latency for hand tracking is described. In one aspect, a method includes accessing a first frame from a camera of an Augmented Reality (AR) device, tracking a first image of a hand in the first frame, rendering virtual content based on the tracking of the first image of the hand in the first frame, accessing a second frame from the camera before the rendering of the virtual content is completed, the second frame immediately following the first frame, tracking, using the computer vision engine of the AR device, a second image of the hand in the second frame, generating an annotation based on tracking the second image of the hand in the second frame, forming an annotated virtual content based on the annotation and the virtual content, and displaying the annotated virtual content in a display of the AR device.

Abstract: Eyewear devices that include two SoCs that share processing workload. Instead of using a single SoC located either on the left or right side of the eyewear devices, the two SoCs have different assigned responsibilities to operate different devices and perform different processes to balance workload. In one example, the eyewear device utilizes a first SoC to operate a first color camera, a second color camera, a first display, and a second display. The first SoC and a second SoC are configured to selectively operate a first and second computer vision (CV) camera algorithms. The first SoC is configured to perform visual odometry (VIO), track hand gestures of the user, and provide depth from stereo images. This configuration provides organized logistics to efficiently operate various features, and balanced power consumption.

Abstract: Eyewear including a frame having a first side and a second side, a first temple extending from the first side of the frame, a second temple extending from the second side of the frame, electronic components, a first system on a chip (SoC) adjacent the first side of the frame coupled to a first set of the electronic components, and a second system on a chip adjacent the second side, the second SoC coupled to the first SoC and to a second set of the plurality of electronic components. Processing workloads are balanced between the first SoC and the second SoC by performing a first set of operations with the first SoC and performing a second set of operations with the second SoC.

Abstract: Examples described herein relate to the dynamic placement of virtual content, such as virtual content widgets, to provide an extended reality (XR) experience. An XR device accesses sensor data from one or more sensors. User context associated with a user of the XR device is determined based on the sensor data. A virtual content widget is identified for presentation to the user. A widget location is selected based on the user context. The XR device causes presentation, via a display component, of the virtual content widget at the widget location.

Abstract: A method for configuring a digital light projector (DLP) of an augmented reality (AR) display device is described. A light source component of the DLP projector is configured to generate a single red-green-blue color sequence repetition per image frame. The AR display device identifies a color sequence of the light source component of the DLP projector and tracks a motion of the AR display device. The AR display device adjusts an operation of the DLP projector based on the single red-green-blue color sequence repetition, the color sequence of the light source component of the DLP projector, and the motion of the AR display device.

Abstract: Eyewear devices that include two SoCs that share processing workload. Instead of using a single SoC located either on the left or right side of the eyewear devices, the two SoCs have different assigned responsibilities to operate different devices and perform different processes to balance workload. In one example, the eyewear device utilizes a first SoC to operate the OS, a first color camera, a second color camera, a first display, and a second display. A second SoC is configured to run computer vision (CV) algorithms, visual odometry (VIO), tracking hand gestures of the user, and providing depth from stereo. This configuration provides organized logistics to efficiently operate various features, and balanced power consumption.

Abstract: Eyewear devices that include two SoCs that share processing workload. Instead of using a single SoC located either on the left or right side of the eyewear devices, the two SoCs have different assigned responsibilities to operate different devices and perform different processes to balance workload. In one example, the eyewear device utilizes a first SoC to operate a first color camera, a second color camera, a first display, and a second display. The first SoC and a second SoC are configured to selectively operate a first and second computer vision (CV) camera algorithms. The first SoC is configured to perform visual odometry (VIO), track hand gestures of the user, and provide depth from stereo images. This configuration provides organized logistics to efficiently operate various features, and balanced power consumption.

Abstract: Eyewear including a frame having a first side and a second side, a first temple extending from the first side of the frame, a second temple extending from the second side of the frame, electronic components, a first system on a chip (SoC) adjacent the first side of the frame coupled to a first set of the electronic components, and a second system on a chip adjacent the second side, the second SoC coupled to the first SoC and to a second set of the plurality of electronic components. Processing workloads are balanced between the first SoC and the second SoC by performing a first set of operations with the first SoC and performing a second set of operations with the second SoC.

Abstract: Examples described herein relate to the dynamic placement of virtual content, such as virtual content widgets, to provide an extended reality (XR) experience. An XR device accesses sensor data from one or more sensors. User context associated with a user of the XR device is determined based on the sensor data. A virtual content widget is identified for presentation to the user. A widget location is selected based on the user context. The XR device causes presentation, via a display component, of the virtual content widget at the widget location.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for continuous surface and depth estimation. A continuous surface and depth estimation system determines the depth and surface normal of physical objects by using stereo vision limited within a predetermined window.

Abstract: Eyewear devices that include two SoCs that share processing workload. Instead of using a single SoC located either on the left or right side of the eyewear devices, the two SoCs have different assigned responsibilities to operate different devices and perform different processes to balance workload. In one example, the eyewear device utilizes a first SoC to operate the OS, a first color camera, a second color camera, a first display, and a second display. A second SoC is configured to run computer vision (CV) algorithms, visual odometry (VIO), tracking hand gestures of the user, and providing depth from stereo. This configuration provides organized logistics to efficiently operate various features, and balanced power consumption.