Video encoding through non-saliency compression for live streaming of high definition videos in low-bandwidth transmission

1. A computer-implemented method of encoding a video stream for high-definition video in a low-bandwidth transmission, the method comprising:

automatically determining a domain including domain-specific characteristics in a high-resolution video stream;

identifying a salient data and a non-salient data in the high-resolution video stream, wherein the identifying of the salient data includes identifying one or more domain-specific characteristics of objects in the video stream;

segmenting the salient data and the non-salient data;

compressing the non-salient data to a lower resolution; and

transmitting the salient data and the compressed non-salient data.

2. The computer-implemented method of claim 1 , further comprising encoding the non-salient data prior to performing the compressing of the non-salient data.

3. The computer-implemented method according to claim 1 , further comprising compressing the salient data at a lower compression ratio than the non-salient data prior to transmitting the salient data and the compressed non-salient data.

4. The computer-implemented method of claim 1 , further comprising identifying at least one of the non-salient data and the salient data in the video stream by a machine learning model.

5. The computer-implemented method of claim 4 , wherein the machine learning model comprises a General Adversarial Network (GAN) machine learning model; and further comprising:

training the GAN machine learning model with data of one or more non-salient features from previously recorded video streams to identify the non-salient data.

6. The computer-implemented method of claim 5 , further comprising providing to a user device one or more of a link to access or a code to execute the GAN machine learning model prior to transmitting the salient data and the compressed non-salient data of the video stream to the user device.

7. The computer-implemented method of claim 1 , wherein the identifying of the salient data includes applying a domain-specific Artificial Intelligence (AI) model for one or more of facial recognition or object recognition.

8. The computer-implemented method of claim 7 , wherein the applying of the domain-specific AI model further comprises identifying a remainder of information of the video stream as the non-salient data.

9. The computer-implemented method of claim 1 , further comprising receiving a plurality of video streams, each video stream having respectively different views of one or more objects, wherein the identifying and segmenting of the salient data and non-salient data is performed individually for at least two respectively different views that are transmitted.

10. A computing device for encoding video streams for high-definition video in a low-bandwidth transmission, the computing device comprising:

a processor;

a memory coupled to the processor, the memory storing instructions to cause the processor to perform acts comprising:

automatically determining a domain including domain-specific characteristics in a high-resolution video stream;

identifying a salient data and a non-salient data in the high-resolution video stream, wherein the identifying of the salient data includes identifying one or more domain-specific characteristics of objects in the video stream;

segmenting the salient data and the non-salient data;

encoding and compressing the non-salient data; and

transmitting the salient data and the compressed non-salient data.

11. The computing device of claim 10 , further comprising:

a General Adversarial Network (GAN) machine learning model in communication with the memory; and

wherein the instructions cause the processor to perform an additional act comprising training the GAN machine learning model with training data of non-salient features based on previously recorded video streams to perform the identifying of at least the non-salient data.

12. The computing device of claim 11 , wherein the instructions cause the processor to perform additional acts comprising:

receiving refined results from the selected agents based on the fused parameters; and

generating a global training model based on the refined results.

13. The computing device of claim 10 , wherein the instructions cause the processor to perform an additional act comprising:

applying a domain-specific Artificial Intelligence (AI) model comprising one or more of facial recognition or object recognition to identify the salient data.

14. The computing device of claim 10 , wherein the instructions cause the processor to perform an additional act comprising transmitting different camera views of the salient data and the non-salient data to a plurality of recipient devices.

15. The computer-implemented method of claim 1 , wherein the compression of the non-salient data is based on transmission of the video stream using the low-bandwidth transmission.