Abstract: Systems, methods and apparatuses are described for managing audio data received from users during a virtual meeting. First and second users, co-located at a particular location, may be participating in the virtual meeting via a first and second computing devices. Third and fourth users, remotely located from the particular location, may be participating in the virtual meeting via one or more computing devices. A first intermediate audio stream corresponding to the detected voice of the third user and a second intermediate audio stream corresponding to the detected voice of the fourth user may be generated. First and second audio output streams may be generated based on the first and second intermediate audio streams. During the virtual meeting, the first audio output stream may be output by the first computing device and the second audio output stream may be output by the second computing device.

Abstract: Methods, systems, and other techniques for providing a biometric reader (e.g., a fingerprint reader) with MEMS components are described herein. An example device includes MEMS components configured to actuate based on detecting contact of a finger with the surface of the device. Control circuitry of the device is configured to detect a finger contacting the surface of the device; identify MEMS components corresponding to the location of the contact; cause the corresponding MEMS components to actuate; measure, using the MEMS components, topography of the finger; detect a fingerprint profile based on the measured topography of the finger; and authenticate the user based at least in part on determining that the measured fingerprint profile corresponds to a registered fingerprint profile.

Abstract: Systems and methods are disclosed for training a model which predicts quality scores for a video subject to a video enhancement technique at various encoding parameters and then utilizing the model to determine whether to apply the video enhancement technique. A model is trained to input a feature vector of a video and output a plurality of output quality scores which indicate an expected quality of the input video subject to a video enhancement technique and encoding parameters. Based on the quality scores, the system determines whether to apply the video enhancement technique for each of the quality parameters.

Abstract: Systems and methods are provided for utilizing historic user consumption of media assets to manage the resources associated with streaming further media asset. tracking, at a computing device, A consumption of a plurality of media assets within a preset time period are tracked at a computing device, and a first resource value associated with the plurality of media assets is identified. A consumption velocity associated with the consumption of the plurality of media assets is identified, and the consumption velocity is weighted based on the first resource value. It is determined that the weighted consumption velocity is above a threshold consumption velocity, and a user interface element comprising a recommendation for a media asset is generated for output, wherein a second resource value associated with the recommended media asset is lower than the first resource value.

Abstract: Systems and methods are described for automatically performing automated radial blurring. A plurality of images may be accessed, and saliency parameters may be determined based on at least one of the plurality of images. Co-saliency parameters may be determined based on the plurality of images. A region of interest (ROI) in the at least one of the plurality of images may be determined based on the saliency parameters and the co-saliency parameters. Automated radial blurring of the at least one of the plurality of images may be performed based on the identified ROI.

Abstract: Disclosed are an immersion liquid supply system and an immersion flow field removal method thereof.

Abstract: Systems and methods are provided herein for providing a personalized selection of modified content items to a user device. Particularly, systems and related methods are disclosed for analyzing content items to determine and store associated dimensions and corresponding sentiments and sentiment intensities and analyzing user behavior to determine and store user preferences comprising a plurality of dimensions and associated user sentiments and user sentiment intensities. Upon receiving a request from a user device for content items about a topic, the system identifies relevant content items, compares the associated dimensions of the identified content items to the user preferences, and modifies the content items accordingly. Following the modification of the content items, the system generates the modified content items for display at the user device.

Abstract: Embodiments herein provide a system for displaying a visual identifier to indicate social dynamics between persons in a subject image. The system analyzes the subject image to detect persons in the subject image. The system identifies the detected persons by searching for an image of a plurality of images that depicts at least one of the detected persons. In some embodiments, the system detects and/or identifies persons based on their facial features. Social dynamics between the detected persons are determined using information corresponding to the plurality of images that depict the detected persons. In some embodiments, the social dynamics provide context for the subject image. In some embodiments, the social dynamics include relationships between the detected persons in the subject image. In some embodiments, the visual identifier is displayed simultaneously with the subject image, such as an overlay on the subject image or a modification to the subject image.

Abstract: Methods, systems, and devices are described herein for providing an extended reality (XR) experience capable of both augmented reality (AR) and virtual reality (VR) functions. The XR device comprises transparent components enabling optical see-through (OST); the device comprising a transparent microLED display and a transparent image sensor together with a meta-lens. The XR OST device can switch between a fully immersive VR mode and an AR mode by controlling the display's transparency. Light from the exterior real-world scene is collected by the meta-lens on a wide field of view (FOV) and focused onto the transparent image sensor to detect objects in the real-world environment.

Abstract: An appearance of a character depicted in a media asset may be altered, for example, by a trained machine learning model, based on demographic attributes of a consumer of the media asset, such as the age or ethnicity of the consumer. A character alteration rule may limit some types of alteration, for example, re-aging a character that would depict a minor in some situation. Altering the appearance of one or more characters depicted in the media asset may also entail altering the appearance of a first character based on demographics of the consumer and automatically altering the appearance of a second character based on the second character's relationship with the first character. The re-aging the second character may be proportional to the re-aging of the first character.

Abstract: Disclosed is an assembly method of a long-span truss, relating to the field of construction engineering. The method includes the following steps: column-top trusses are pre-assembled at column tops of structural columns. Middle trusses are assembled on platforms between adjacent structural columns. Column-top hydraulic lifting devices are installed on the column-top trusses. Hoisting points are provided at both ends of the middle truss. The column-top hydraulic lifting devices are connected to corresponding hoisting points. The middle truss is lifted to a design height. The middle truss is connected to the column-top truss. The column-top hydraulic lifting devices are dismantled.

Abstract: Systems and methods are provided for selectively storing and retrieving video data by converting video segments into textual descriptions associated with key visual features and incrementally building a vector database of feature embeddings. The feature embeddings can be used in combination with text-to-video generation models to reconstruct video segments on demand. Selective storage of feature embeddings enables optimized, accurate, and contextually relevant text-to-text and text-to-video generation and video searching functionalities.

Abstract: Systems and methods are disclosed for enhancing output of deemphasized content. A content item may be provided for output on a device. One or more attributes associated with text and/or audio in at least one segment of a content item may be identified. The attributes of the text and/or audio may not be present in other segments of the content item. Based on determining that the attributes of the text and/or audio indicate an intent to deemphasize the text and/or audio, the output of the text and/or audio on the device is enhanced.

Abstract: Systems and methods for displaying a processed representation of media content on an extended reality (XR) device are disclosed herein. The XR device identifies a media device in its physical environment displaying media content and captures a pass-through video of the environment, including a representation of the media device showing the content. The representation is defined, in part, by a display parameter with a first value. The XR device establishes a wireless connection with the media device to receive data for displaying a representation of the content with a display parameter having a second value. Based on its display capabilities and the received data, the XR device processes the representation to adjust the display parameter to a second value, enhancing the content presentation. The processed representation is then displayed on the XR device, ensuring the media content aligns with the XR device's capabilities for an optimized user experience.

Abstract: Systems and methods are provided for recommending voice-based commands based on preferred speech characteristics. A voice-based interface receives a voice input including a first command. The voice-based interface determines, based on one or more user preferences, that the command does not fulfill preferred speech characteristics indicative of a speech efficiency and/or a speech naturalness. The voice-based interface identifies a second command that fulfills the preferred speech characteristics indicative of one or both of the speech efficiency and the speech naturalness. The voice-based interface generates for output a response including a recommendation of the second command.

Abstract: The present application relates to the technical field of LED display, and discloses an LED module having through channels and conforming to certification standards of DCI and a display screen.

Abstract: Systems and methods are provided for dynamically adjusting picture resolution in streaming video in response to estimated bandwidth changes in the network connection. During encoding of a first video segment of a source video, a change in the estimated bandwidth may be detected. To account for changes in the estimated bandwidth, first image frame quality data based on first segment image frames in the first video segment and second image frame quality data based on a subsequent image frame in a second video segment of the source video are generated. The bitrate and picture resolution for the second video segment are determined based on the change in the estimated bandwidth and a comparison of the first image frame quality data with the second image frame quality data. The second video segment may then be encoded for streaming at the bitrate and picture resolution determined by the comparison.

Abstract: The present disclosure is related to systems and methods for synchronizing content items with future events, for example, by dynamically modifying content item playlists. The systems and methods may identify a playlist and a future time at which a time-sensitive content item is to be played. The systems and methods may determine an estimated time of playback of the time-sensitive content item in the playlist and may modify, based on detecting a difference between the estimated time of playback and the identified future time, the playlist to achieve playback of the time-sensitive content item at the identified future time. The systems and methods may modify the playlist by modifying one or more durations of the content items in the playlist based on a plurality of modifiability scores.

Abstract: The present disclosure relates to methods and systems for improving, by a device, user engagement on a media content item being presented. To do so, the device presents, to the user, a poll related to the media content item during the presentation of the media content item. The device generates the poll by e.g., analyzing the media content item to identify multiple events, categorizing the multiple events based on the analyzing, selecting an event category and selecting randomly at least one event outcome within the selected event category. The device is able to choose, upon a user input, an event outcome among the randomly-selected at least one event outcome.

Abstract: The present disclosure relates to methods and systems for e.g., eliminating, in a live capturing by a camera during a video conference, a reflection of an image rendered by a computer screen, on eyeglasses worn by a user facing said computer screen. This reflection tends to divert the other attendees' attention, which may extend the duration of the video conference. The computer screen alternates between a regular display state and a black-frame display state while the camera advantageously captures a series of images depicting the user's face only when the computer screen is in the black-frame display state, during which the computer screen does not emit any light. Each image of the captured series of images accordingly depicts the user's face wearing eyeglasses through which the user's eyes clearly appear unobstructed by the aforementioned reflection.