公开(公告)号：US11871068B1

公开(公告)日：2024-01-09

申请号：US16712470

申请日：2019-12-12

Applicant: Amazon Technologies, Inc.

Inventor： Christian Garcia Siagian ,  Ryan Barlow Dall ,  Charles Effinger ,  Ramakanth Mudumba

IPC: H04N21/43 ,  H04N5/06 ,  G06V20/40 ,  G06V40/16 ,  H04N21/233

CPC classification number: H04N21/4307 ,  G06V20/46 ,  G06V40/165 ,  G06V40/171 ,  H04N5/06 ,  H04N21/233 ,  H04N21/4302

Abstract: Techniques for identifying synchronization errors between audio and video are described herein. Audio portions in audio for media content may be identified based at least in part on a sound level associated with first respective segments of the audio portions. A subset of the audio portions may be selected based at least in part on a duration associated with the audio portions. For a segment of the subset a first number of frames in the audio and a second number of frames in the video for the segment may be determined. A determination may be made that the segment includes a conversation segment based at least in part on the first number of frames, the second number of frames, and a first threshold. A synchronization error may be identified in the conversation segment based on a difference between the audio and the video of the conversation segment.

公开(公告)号：US10904476B1

公开(公告)日：2021-01-26

申请号：US16712294

申请日：2019-12-12

Applicant: Amazon Technologies, Inc.

Inventor： Christian Garcia Siagian ,  Charles Effinger ,  David Niu ,  Yang Yu ,  Narayan Sundaram ,  Arjun Cholkar ,  Ramakanth Mudumba

IPC: H04N7/01 ,  G06K9/00 ,  G06T3/40

Abstract: Techniques for automated up-sampling of media files are provided. In some examples, a title associated with a media file, a metadata file associated with the title, and the media file may be received. The media file may be partitioned into one or more scene files, each scene file including a plurality of frame images in a sequence. One or more up-sampled scene files may be generated, each corresponding to a scene file of the one or more scene files. An up-sampled media file may be generated by combining at least a subset of the one or more up-sampled scene files. Generating one or more up-sampled scene files may include identifying one or more characters in a frame image of the plurality of frame images, based at least in part on implementation of a facial recognition algorithm including deep learning features in a neural network.

公开(公告)号：US12190871B1

公开(公告)日：2025-01-07

申请号：US17468415

申请日：2021-09-07

Applicant: Amazon Technologies, Inc.

Inventor： Christian Garcia Siagian ,  Charles Effinger ,  Nicholas Ren-Jie Capel ,  Jobel Kyle Petallana Vecino ,  Gordon Zheng ,  Kymry Michael Burwell ,  Stephen Andrew Low

IPC: G10L15/04 ,  G06Q30/0241 ,  G10L15/16 ,  G10L15/18

Abstract: Techniques and methods are disclosed for detecting long-form audio content in one or more audio files. A computing system receives first audio data corresponding to a first version of an audio file and second audio data corresponding to a second version of the audio file. The computing system generates a first transcript of the first audio data and a second transcript of the second audio data. The computing system compares the first audio data and the second audio data and the first transcript and the second transcript to identify advertisement portions and content portions of the audio data. Using a semantic model based on a machine learning (ML) transformer, the computing system can determine advertisement segments within the advertisement portions, the advertisement segments corresponding to separate advertisements. Information corresponding to the duration and location of the advertisement segments is stored in a data store of the computing system.

公开(公告)号：US11070891B1

公开(公告)日：2021-07-20

申请号：US16708996

申请日：2019-12-10

Applicant: Amazon Technologies, Inc.

Inventor： Charles Effinger ,  Ryan Barlow Dall ,  Christian Garcia Siagian ,  Ramakanth Mudumba ,  Lawrence Kyuil Chang

IPC: H04N21/488 ,  H04N21/43 ,  H04N21/442 ,  H04N21/4223 ,  G10L15/26

Abstract: A subtitle management system is provided that analyzes and adjusts subtitles for video content to improve the experience of viewers. Subtitles may be optimized or otherwise adjusted to display in particular regions of the video content, to display in synchronization with audio presentation of the spoken dialogue represented by the subtitles, to display in particular colors, and the like. Subtitles that are permanently integrated into the video content may be identified and addressed. These and other adjustments may be applied to address any of a variety of subtitle issues and shortcomings with conventional methods of generating subtitles.