TECHNIQUES FOR TRAINING A MULTITASK LEARNING MODEL TO ASSESS PERCEIVED AUDIO QUALITY

    公开(公告)号:US20210350819A1

    公开(公告)日:2021-11-11

    申请号:US16905793

    申请日:2020-06-18

    Applicant: NETFLIX, INC.

    Abstract: In various embodiments, a training application trains a multitask learning model to assess perceived audio quality. The training application computes a set of pseudo labels based on a first audio clip and multiple models. The set of pseudo labels specifies metric values for a set of metrics that are relevant to audio quality. The training application also computes a set of feature values for a set of audio features based on the first audio clip. The training application trains a multitask learning model based on the set of feature values and the set of pseudo labels to generate a trained multitask learning model. In operation, the trained multitask learning model maps different sets of feature values for the set of audio features to different sets of predicted labels. Each set of predicted labels specifies estimated metric values for the set of metrics.

    TECHNIQUES FOR DETECTING CHANGES TO CIRCUIT DELAYS IN TELECOMMUNICATIONS NETWORKS

    公开(公告)号:US20210306245A1

    公开(公告)日:2021-09-30

    申请号:US16836505

    申请日:2020-03-31

    Applicant: NETFLIX, INC.

    Abstract: In various embodiments, a monitoring application assesses delays associated with a circuit within a network. The monitoring application determines a measured trip time between a first device and a second device that is connected to the first device via the circuit. The measured trip time is associated with a first variance attributable to the first device. The monitoring application performs one or more digital signal processing operations based on the measured trip time to generate a predicted trip time. The predicted trip time is associated with a second variance attributable to the first device that is less than the first variance. Based on the predicted trip time, the monitoring application determines characteristic(s) of the delay associated with the circuit. Advantageously, reducing variations attributable to the first device when generating the first predicted trip time increases the accuracy with which the monitoring application can determine the characteristic(s) of the delay.

    Dynamically adjusting text strings based on machine translation feedback

    公开(公告)号:US11126799B2

    公开(公告)日:2021-09-21

    申请号:US16289999

    申请日:2019-03-01

    Applicant: Netflix, Inc.

    Abstract: The disclosed computer-implemented method may include accessing a string of text that includes characters written in a first language. The method may next include translating the text string into different languages using machine translation. The method may next include identifying, among the translated text strings, a shortest string and a longest string. The method may also include calculating a customized string length adjustment ratio for adjusting the length of the accessed text string based on the shortest translated string length and the longest translated string length. Furthermore, the method may include dynamically applying the calculated customized string length adjustment ratio to the accessed text string, so that the length of the accessed text string may be dynamically adjusted according to the customized string length adjustment ratio. The method may also include presenting the adjusted text string in the user interface. Various other methods, systems, and computer-readable media are also disclosed.

    TEMPORAL PLACEMENT OF A REBUFFERING EVENT

    公开(公告)号:US20210289258A1

    公开(公告)日:2021-09-16

    申请号:US17243389

    申请日:2021-04-28

    Applicant: Netflix, Inc.

    Abstract: A method includes receiving, with a computing system, data representing a video item into a buffer. The method further includes outputting the video item from the buffer to a display system. The method further includes determining that utilization of the buffer falls below a predetermined threshold. The method further includes, in response to determining that the utilization of the buffer falls below the predetermined threshold, determining that there is a specified rebuffering point within a predetermined time frame. The method further includes pausing with the computing system, the video item at the specified rebuffering point in response to determining that there is the specified rebuffering point within the predetermined time frame.

    TECHNIQUES FOR EFFICIENTLY PERFORMING SUBSEQUENCE-BASED ENCODING FOR A MEDIA TITLE

    公开(公告)号:US20210281894A1

    公开(公告)日:2021-09-09

    申请号:US17328358

    申请日:2021-05-24

    Applicant: NETFLIX, INC.

    Abstract: In various embodiments, an interpolation-based encoding application encodes a first subsequence included in a media title at each encoding point included in a first set of encoding points to generate encoded subsequences. Subsequently, the interpolation-based encoding application performs interpolation operation(s) based on the encoded subsequences to estimate a first media metric value associated with a first encoding point that is not included in the first set of encoding points. The interpolation-based encoding application then generates an encoding recipe based on the encoded subsequences and the first media metric value. The encoding recipe specifies a different encoding point for each subsequence included in the media title. After determining that the encoding recipe specifies the first encoding point for the first subsequence, the interpolation-based encoding application encodes the first subsequence at the first encoding point to generate at least a portion of an encoded version of the media title.

    Global approach to buffering media content

    公开(公告)号:US11115688B2

    公开(公告)日:2021-09-07

    申请号:US16716405

    申请日:2019-12-16

    Applicant: NETFLIX, INC.

    Inventor: Mark Watson

    Abstract: In various embodiments, a graph streaming application computes a first distance along a first potential playback path between a first playback position and a first media content block. The graph streaming application then computes a first score for the first media content block based on the first distance and a first probability associated with the first potential playback path. Similarly, the graph streaming application computes a second score for a second media content block based on a second playback position, a second potential playback path, and a second probability associated with the second potential playback path. The graph streaming application compares the first score and the second score to determine that the first media content block should be buffered by a client device. The graph streaming application then causes the first media content block to be stored in a playback buffer for subsequent playback on the client device.

    Techniques for detecting spatial anomalies in video content

    公开(公告)号:US11107206B2

    公开(公告)日:2021-08-31

    申请号:US16143386

    申请日:2018-09-26

    Applicant: NETFLIX, INC.

    Abstract: In various embodiments, a defective pixel detection application automatically detects defective pixels in video content. In operation, the defective pixel detection application computes a first set of pixel intensity gradients based on a first frame of video content and a first neighborhood of pixels associated with a first pixel. The defective pixel detection application also computes a second set of pixel intensity gradients based on the first frame and a second neighborhood of pixels associated with the first pixel. Subsequently, the defective pixel detection application computes a statistical distance between the first set of pixel intensity gradients and the second set of pixel intensity gradients. The defective pixel detection application then determines that the first pixel is defective based on the statistical distance.

    FRAME RATE CONVERSION
    79.
    发明申请

    公开(公告)号:US20210250545A1

    公开(公告)日:2021-08-12

    申请号:US17244943

    申请日:2021-04-29

    Applicant: Netflix, Inc.

    Abstract: The disclosed method includes accessing video content encoded at a specified frame rate, and determining a refresh rate for an electronic display on which the video content is to be presented. The method next includes specifying a time interval for the video content over which frame rate conversion is to be applied to synchronize the video content frame rate with the electronic display refresh rate. The method also includes presenting the video content on the electronic display where the playback speed is adjusted for a first part of the interval. At this adjusted speed, the interval is played back using original video frames and multiple frame duplications. The presenting also adjusts playback speed of a second part of the interval. At the adjusted speed, the interval is played back using the original frames and a different number of frame duplications. Various other methods, systems, and computer-readable media are also disclosed.

    Extension for targeted invalidation of cached assets

    公开(公告)号:US11086960B2

    公开(公告)日:2021-08-10

    申请号:US16926555

    申请日:2020-07-10

    Applicant: NETFLIX, INC.

    Abstract: Various embodiments of the present disclosure relate to a computer-implemented method of receiving an invalidation command that includes a first asset identifier, and a set of extension value fields, identifying a first file that is stored in a first cache device and is associated with an identifier matching the first asset identifier, comparing attributes of the first file to the set of extension value fields, and upon determining that, for each extension value field, at least one attribute of the first file matches at least one value of the extension value field, causing the first file in the first cache device to be preserved, or upon determining that, for at least one extension value, at least one attribute of the first file does not match at least one value of the extension value field, causing the first file to be invalidated.

Patent Agency Ranking