MEDIA IDENTIFIER GENERATION FOR CAMERA-CAPTURED MEDIA

    公开(公告)号:US20190043534A1

    公开(公告)日:2019-02-07

    申请号:US16153304

    申请日:2018-10-05

    申请人: GoPro, Inc.

    摘要: A video identifier uniquely identifying a video captured by a camera is generated. The video includes video frames and optionally concurrently captured audio as well as video metadata describing the video. Video data is extracted from at least two of the video's frames. By combining the extracted video data in an order specified by an identifier generation protocol, an extracted data object is generated. The extracted data object is hashed to generate the unique media identifier, which is stored in association with the video. The identifier generation protocol may indicate the portions of the video data to extract, such as video data corresponding to particular video frames and audio data corresponding to particular audio samples. The extracted data object may include a size of particular video frames, a number of audio samples in the video, or the duration of the video, for example.

    Signaling random access points for streaming video data

    公开(公告)号:US09992555B2

    公开(公告)日:2018-06-05

    申请号:US14741057

    申请日:2015-06-16

    摘要: This disclosure describes techniques relevant to HTTP streaming of media data. According to these techniques, a server device may signal an open decoding refresh (ODR) random access point (RAP) for a movie segmentation of a movie representation. At least one frame of the media segmentation following the ODR RAP frame in decoding order may not be correctly decoded, and wherein each frame of the media segmentation following the ODR RAP frame in display order can be correctly decoded without relying on content of frames prior to the ODR RAP in display order. According to the techniques of this disclosure, a client device may communicate a request to a server device for the streaming of media data based on signaling of the ODR RAP. Also according to the techniques of this disclosure, a client device may commence decoding and/or playback of the movie representation based on signaling of the ODR RAP.