Video tagging for video communications

1. A method for video communications, comprising:

transmitting, from a first apparatus using a network, a first video stream of a first user to a second apparatus of a second user,

wherein the second user is in video communication with the first user;

receiving, from the second apparatus using the network, a second video stream of the second user;

determining, using the second video stream by a processor, a reaction of the second user to the first video stream,

wherein the reaction is indicative of a low quality of the first video stream;

associating the reaction of the second user with the first video stream, wherein the associating the reaction of the second user with the first video stream comprises:

determining, based on a reaction timestamp of the reaction associated with a decoded picture of the second video stream, a portion of the first video stream;

determining an encoding timestamp of the first apparatus of the portion of the first video stream; and

associating, in a database, the reaction of the second user with the portion of the first video stream using the reaction timestamp and the encoding timestamp; and

updating, in response to the reaction of the second user, a parameter for encoding the first video stream at the first apparatus so as to increase a quality of the first video stream.

2. The method of claim 1 , wherein the reaction of the second user to the first video stream is determined by one of the first apparatus, a server in the network connected to the first apparatus, the second apparatus, or a server in the network connected to the second apparatus.

3. The method of claim 1 , wherein determining the reaction of the second user to the first video stream comprises:

determining the reaction of the second user by analyzing at least one of a decoded picture selected from decoded pictures of the second video stream, or an audio decoded from the second video stream.

4. The method of claim 3 , wherein a temporal difference between the reaction timestamp and the encoding timestamp is less than a predetermined threshold.

5. The method of claim 3 , wherein

the reaction of the second user is determined based on at least one of a facial expression of the second user, an emotion of the second user, a gesture of the second user, a characteristic of a speech of the second user, and a semantic meaning of the speech of the second user,

the facial expression, the emotion, and the gesture are determined by analyzing the decoded picture, and

the speech and the semantic meaning of the speech are determined by analyzing the audio.

6. The method of claim 5 , wherein

the facial expression and the emotion of the second user are determined from decoded pictures of the second video stream using one of a convolutional neural network (CNN) technique, a k-nearest neighbors (k-NN) technique, and a support vector machine (SVM) technique,

the characteristic of the speech of the second user is determined from an audio associated with the second video stream, and

the semantic meaning of the speech of the second user is determined from the audio using an automatic speech recognition (ASR) technique.

7. The method of claim 1 , wherein the parameter comprises at least one of a bitrate, a frame rate, a resolution, or a forward error correction (FEC) ratio.

8. An apparatus for video communications, comprising:

a processor; and

a memory coupled to the processor, the memory configured to store instructions which when executed by the processor become operational with the processor to:

transmit, from a first user apparatus using a network, a first video stream of a first user to a second user apparatus of a second user, wherein the second user is in video communication with the first user;

receive, from the second user apparatus using the network, a second video stream of the second user;

determine, using the second video stream, a reaction of the second user to the first video stream, wherein the reaction of the second user is indicative of a quality of the first video stream;

determining, based on a reaction timestamp of the reaction associated with a decoded picture of the second video stream, a portion of the first video stream;

determine an encoding timestamp of the first apparatus of the portion of the first video stream; and

associating the reaction of the second user with the portion of the first video stream using the reaction timestamp and the encoding timestamp; and

in response to the reaction indicating a low quality of the first video stream, update a parameter of encoding the first video stream at the first user apparatus so as to increase the quality of the first video stream.

9. The apparatus of claim 8 , wherein the memory comprising the instructions operational with the processor to determine the reaction of the second user to the first video stream further comprises instructions which when executed by the processor become operational with the processor to:

determine, using the second video stream, multiple reactions of the second user to respective portions of the first video stream.

10. The apparatus of claim 8 , wherein the memory comprising the instructions operational with the processor to determine the reaction of the second user to the first video stream further comprises instructions which when executed by the processor become operational with the processor to:

determine the reaction of the second user by analyzing at least one of a decoded picture selected from decoded pictures of the second video stream, and an audio decoded from the second video stream.

11. The apparatus of claim 10 , wherein a temporal difference between the reaction timestamp and the encoding timestamp is less than a predetermined threshold.

12. The apparatus of claim 10 , wherein

the reaction of the second user is determined based on at least one of a facial expression of the second user, an emotion of the second user, a gesture of the second user, a characteristic of a speech of the second user, and a semantic meaning of the speech of the second user,

the facial expression, the emotion, and the gesture are determined by analyzing the decoded picture, and

the speech and the semantic meaning of the speech are determined by analyzing the audio.

13. The apparatus of claim 12 , wherein

the facial expression and the emotion of the second user are determined from decoded pictures of the second video stream using one of a convolutional neural network (CNN) technique, a k-nearest neighbors (k-NN) technique, and a support vector machine (SVM) technique,

the characteristic of the speech of the second user is determined from an audio associated with the second video stream, and

the semantic meaning of the speech of the second user is determined from the audio using an automatic speech recognition (ASR) technique.

14. The apparatus of claim 8 , wherein the parameter comprises at least one of a bitrate, a frame rate, a resolution, and a forward error correction (FEC) ratio.

15. A system for tagging a video for video communications, comprising:

a first apparatus in a network, the first apparatus associated with a first user;

a second apparatus in the network, the second apparatus associated with a second user, wherein the second user is in video communication with the first user; and

a server in the network connected to the first apparatus, the server comprising a processor and a memory coupled to the processor, the memory configured to store instructions which when executed by the processor become operational with the processor to:

transmit, using the network, a first video stream of the first user from the first apparatus to the second apparatus;

receive, from the second apparatus using the network, a second video stream of the second user;

determine, using the second video stream, a reaction of the second user to the first video stream, wherein the reaction of the second user is indicative of a quality of the first video stream;

determine, based on a reaction timestamp of the reaction associated with a decoded picture of the second video stream, a portion of the first video stream;

determine an encoding timestamp of the first apparatus of the portion of the first video stream; and

associate the reaction of the second user with the portion of the first video stream using the reaction timestamp and the encoding timestamp; and

in response to the reaction indicative a low quality of the first video stream, update a parameter of encoding the first video stream at the first apparatus so as to increase the quality of the first video stream.

16. The system of claim 15 , wherein the memory comprising the instructions operational with the processor to determine the reaction of the second user to the first video stream further comprises instructions which when executed by the processor become operational with the processor to:

determine, using the second video stream, multiple reactions of the second user to respective portions of the first video stream.

17. The system of claim 15 , wherein the memory comprising the instructions operational with the processor to determine the reaction of the second user to the first video stream further comprises instructions which when executed by the processor become operational with the processor to:

determine the reaction of the second user by analyzing at least one of a decoded picture selected from decoded pictures of the second video stream, and an audio decoded from the second video stream, wherein

the reaction of the second user is determined based on at least one of a facial expression of the second user, an emotion of the second user, a gesture of the second user, a characteristic of a speech of the second user, and a semantic meaning of the speech of the second user,

the facial expression, the emotion, and the gesture are determined by analyzing the decoded picture, and

the speech and the semantic meaning of the speech are determined by analyzing the audio.

18. The system of claim 17 , wherein a temporal difference between the reaction timestamp and the encoding timestamp is less than a predetermined threshold.

19. The system of claim 17 , wherein

the facial expression and the emotion of the second user are determined from decoded pictures of the second video stream using one of a convolutional neural network (CNN) technique, a k-nearest neighbors (k-NN) technique, and a support vector machine (SVM) technique,

the characteristic of the speech of the second user is determined from an audio associated with the second video stream, and

the semantic meaning of the speech of the second user is determined from the audio using an automatic speech recognition (ASR) technique.

20. The system of claim 15 , wherein the parameter comprises at least one of a bitrate, a frame rate, a resolution, and a forward error correction (FEC) ratio.