Artificial latency for moderating voice communication

1. A computer-implemented method to determine whether to introduce latency into an audio stream from a particular speaker, the method comprising:

receiving an audio stream from a sender device;

providing, as input to a trained machine-learning model, the audio stream and a speech analysis score, information about one or more voice emotion parameters, and one or more voice emotion scores for a first user associated with the sender device, wherein the trained machine-learning model is iteratively applied to the audio stream and wherein each iteration corresponds to a respective portion of the audio stream;

generating as output, with the trained machine-learning model, a level of toxicity in the audio stream;

identifying silence or a pause between words in the audio stream, the silence or the pause corresponding to a particular timestamp in the audio stream; and

transmitting the audio stream to a recipient device, wherein the transmitting is performed to introduce a time delay in the audio stream based on the level of toxicity and wherein the time delay is introduced as a gap in the audio stream at the particular timestamp of the silence or the pause between words.

2. The method of claim 1 , further comprising:

identifying an instance of toxicity in the audio stream; and

replacing the instance of toxicity in the audio stream with a noise or silence before transmitting the audio stream to the recipient device.

3. The method of claim 1 , wherein the audio stream is part of a metaverse.

4. The method of claim 2 , further comprising:

updating the speech analysis score based on identifying the instance of toxicity in the audio stream.

5. The method of claim 1 , further comprising:

receiving text from a text channel associated with the sender device, wherein the text channel is separate from the audio stream; and

generating a text score that indicates a toxicity rating for the text;

wherein the input to the trained machine-learning model further includes the text score.

6. The method of claim 1 , wherein the input to the trained machine-learning model further includes a toxicity history of the first user, a speaker history and metadata associated with the first user, and a listener history and metadata associated with a second user associated with the recipient device.

7. The method of claim 1 , wherein the one or more voice emotion parameters include tone, pitch, and a vocal effort level that are determined based on one or more prior audio streams from the sender device.

8. The method of claim 1 , wherein the audio stream is provided with a visual signal, and the method further comprises synchronizing the visual signal to the audio stream by introduction of a same time delay in the visual signal as the time delay of the audio stream.

9. The method of claim 1 , wherein the audio stream is part of a video stream, and the method further comprises:

analyzing the audio stream to identify an instance of toxicity;

responsive to identifying the instance of toxicity, detecting a portion of the video stream that depicts an offensive action, wherein the offensive action occurs within a predetermined time period of the instance of toxicity; and

responsive to detecting the offensive action, modifying at least the portion of the video stream by one or more of blurring the portion or replacing the portion with pixels that match a background region.

10. The method of claim 1 , wherein the audio stream is part of a video stream, and the method further comprises:

performing motion detection on the video stream to detect an offensive gesture; and

responsive to detecting the offensive gesture, modifying at least the portion of the video stream by one or more of blurring the portion or replacing the portion with pixels that match a background region.

11. The method of claim 1 , wherein if the level of toxicity is below a minimum threshold, the time delay is zero seconds.

12. A device comprising:

a processor; and

a memory coupled to the processor, with instructions stored thereon that, when executed by the processor, cause the processor to perform operations comprising:

receiving an audio stream from a sender device;

providing, as input to a trained machine-learning model, the audio stream and a speech analysis score, information about one or more voice emotion parameters, and one or more voice emotion scores for a first user associated with the sender device, wherein the trained machine-learning model is iteratively applied to the audio stream and wherein each iteration corresponds to a respective portion of the audio stream;

generating as output, with the trained machine-learning model, a level of toxicity in the audio stream;

identifying silence or a pause between words in the audio stream, the silence or the pause corresponding to a particular timestamp in the audio stream; and

transmitting the audio stream to a recipient device, wherein the transmitting is performed to introduce a time delay in the audio stream based on the level of toxicity and wherein the time delay is introduced as a gap in the audio stream at the particular timestamp of the silence or the pause between words.

13. The device of claim 12 , wherein the operations further comprise:

identifying an instance of toxicity in the audio stream; and

replacing the instance of toxicity in the audio stream with a noise or silence before transmitting the audio stream to the recipient device.

14. The device of claim 12 , wherein the audio stream is part of a metaverse.

15. The device of claim 12 , wherein the operations further comprise:

updating the speech analysis score based on identifying the instance of toxicity in the audio stream.

16. A non-transitory computer-readable medium with instructions stored thereon that, when executed by one or more computers, cause the one or more computers to perform operations, the operations comprising:

receiving an audio stream from a sender device;

providing, as input to a trained machine-learning model, the audio stream and a speech analysis score, information about one or more voice emotion parameters, and one or more voice emotion scores for a first user associated with the sender device, wherein the trained machine-learning model is iteratively applied to the audio stream and wherein each iteration corresponds to a respective portion of the audio stream;

generating as output, with the trained machine-learning model, a level of toxicity in the audio stream;

identifying silence or a pause between words in the audio stream, the silence or the pause corresponding to a particular timestamp in the audio stream; and

transmitting the audio stream to a recipient device, wherein the transmitting is performed to introduce a time delay in the audio stream based on the level of toxicity and wherein the time delay is introduced as a gap in the audio stream at the particular timestamp of the silence or the pause between words.

17. The computer-readable medium of claim 16 , wherein the operations further comprise:

identifying an instance of toxicity in the audio stream; and

replacing the instance of toxicity in the audio stream with a noise or silence before transmitting the audio stream to the recipient device.

18. The computer-readable medium of claim 16 , wherein the audio stream is part of a metaverse.

19. The computer-readable medium of claim 16 , wherein the operations further comprise:

updating the speech analysis score based on identifying the instance of toxicity in the audio stream.

20. The computer-readable medium of claim 16 , wherein the operations further comprise:

receiving text from a text channel associated with the sender device, wherein the text channel is separate from the audio stream; and

generating a text score that indicates a toxicity rating for the text;

wherein the input to the trained machine-learning model further includes the text score.