Method and apparatus for transmitting adaptive video in real time using content-aware neural network
A method and apparatus for transmitting adaptive video in real time using a content-aware neural network are disclosed. At least one embodiment provides a method performed by a server for transmitting an adaptive video in real time by using content-aware deep neural networks (DNNs), including downloading a video, encoding a downloaded video for each of at least one resolution, dividing an encoded video into video chunks of a predetermined size, training the content-aware DNNs by using encoded video, generating a configuration or manifest file containing information on trained content-aware DNNs and information on the encoded video, and transmitting the configuration file upon a request of a client.
1. A method performed by a server for transmitting an adaptive video in real time by using content-aware deep neural networks (DNNs), the method comprising:
downloading a video;
encoding a downloaded video for each of at least one resolution;
dividing an encoded video into video chunks of a predetermined size;
training the content-aware DNNs by using the encoded video;
generating a configuration file containing information on trained content-aware DNNs and information on the encoded video; and
transmitting the configuration file upon a request of a client.
2. The method of claim 1 , wherein the configuration file comprises information about a storage location of the encoded video.
3. The method of claim 1 , wherein the configuration file comprises information about at least one of a layer count, a channel count, storage locations, sizes, and degrees of quality enhancement of the trained content-aware DNNs.
4. The method of claim 1 , wherein the content-aware DNNs each comprises a requisite component that is necessarily executed and an optional component that is optionally executed.
5. A server apparatus for transmitting an adaptive video in real time by using content-aware deep neural networks (DNNs), the server apparatus comprising:
a transmission and reception unit configured to download a video and to transmit a configuration file at a client's request; and
a control unit configured
to encode a downloaded video for each of at least one resolution,
to divide an encoded video into video chunks of a predetermined size,
to train the content-aware DNNs by using the encoded video, and
to generate the configuration file containing information on trained content-aware DNNs and information on the encoded video.
6. The server apparatus of claim 5 , wherein the information on the encoded video includes at least one of storage locations, resolutions, and bitrates of the encoded video.
7. The server apparatus of claim 5 , wherein the information on the trained content-aware DNNs includes at least one of a layer count, a channel count, storage locations, sizes, and degrees of quality enhancement of the trained content-aware DNNs.
8. The server apparatus of claim 1 , wherein the content-aware DNNs each comprises a requisite component that is necessarily executed and an optional component that is optionally executed.
9. The method of claim 1 , wherein the training the content-aware DNNs comprises training the content-aware DNNs by processing the encoded video through a plurality of different processing paths, and
wherein the plurality of different processing paths comprises a first processing path that includes only requisite neural network layers, and a second image processing path that includes redundant layers in addition to the requisite neural network layers.