公开(公告)号：US20210390315A1

公开(公告)日：2021-12-16

申请号：US17344752

申请日：2021-06-10

Applicant: NETFLIX, INC.

Inventor： Shervin Ardeshir BEHROSTAGHI ,  Nagendra K. KAMATH

IPC: G06K9/00

Abstract: One embodiment of the present invention sets forth a technique for selecting a frame of video content that is representative of a media title. The technique includes applying an embedding model to a plurality of faces included in a set of frames of the video content to generate a plurality of face embeddings. The technique also includes aggregating the plurality of face embeddings into a plurality of clusters representing a plurality of characters included in the media title. The technique further includes computing a plurality of prominence scores for the plurality of characters based on one or more attributes of the plurality of clusters, and selecting, from the set of frames, a frame of video content as representative of the media title based on one or more prominence scores for one or more characters included in the frame.

公开(公告)号：US20250045633A1

公开(公告)日：2025-02-06

申请号：US18749378

申请日：2024-06-20

Applicant: NETFLIX, INC.

Inventor： Shervin Ardeshir BEHROSTAGHI ,  Qi QI

IPC: G06N20/00

Abstract: In various embodiments, a model trainer application trains a machine learning model with improved identity robustness. The model trainer application first processes images of faces using a trained face recognition model to generate a proxy representation of an identity of the individual in each image. Representations of individuals with similar faces lie in the same neighborhoods within a proxy identity space. The model trainer application trains a machine learning model to perform a task relating to faces while considering the accuracy of each identity proxy neighborhood. The model trainer assigns different weights to each image sample in a neighborhood based on the number of samples with the same output class in that neighborhood. The assigned weights can then be used to compute a relatively unbiased identity loss function that is used to train the machine learning model to perform the task relating to faces while being robust to identity features.

公开(公告)号：US20240242501A1

公开(公告)日：2024-07-18

申请号：US18620764

申请日：2024-03-28

Applicant: NETFLIX, INC.

Inventor： Shervin Ardeshir BEHROSTAGHI ,  Nagendra K. KAMATH

IPC: G06V20/40 ,  G06V40/16

CPC classification number: G06V20/47 ,  G06V20/49 ,  G06V40/172

Abstract: One embodiment of the present invention sets forth a technique for selecting a frame of video content that is representative of a media title. The technique includes applying an embedding model to a plurality of faces included in a set of frames of the video content to generate a plurality of face embeddings. The technique also includes aggregating the plurality of face embeddings into a plurality of clusters representing a plurality of characters included in the media title. The technique further includes computing a plurality of prominence scores for the plurality of characters based on one or more attributes of the plurality of clusters, and selecting, from the set of frames, a frame of video content as representative of the media title based on one or more prominence scores for one or more characters included in the frame.