Automatic speaker identification
Simon Says automatically identifies speakers in your transcript; simply label one row and watch other rows be labeled too.
Note: To circumvent auto-speaker identification, watch this video.
Simon Says A.I. automatically identifies speakers in a transcript for many languages.
Simply label a row with a speaker name.
And see other rows get labeled with that same speaker name where Simon Says finds a match.
To add speaker labels, click the > to the left of a row, click in the "Add speaker" box, and type in the name.
When you add a speaker label, the A.I. will try to find other rows with the same speaker and automatically label them for you.
The AI tries to match based on traits like tone of voice. Speaker identification A.I. is hard because people change their tone even if the same sentence! To see the best results in accuracy, record the audio in high quality. This means ideally a mic per speaker, one person speaking at a time, and with low/no background noise. (Also the more pronounced the difference between speakers, the better the accuracy as it is easier to distinguish between speakers.)
If the wrong speaker is labeled on any row, change it by selecting a different speaker label for that row, adding a new label, or selecting "No speaker".
Don't want all the speaker labels tagged?
Delete the first instance of the label by clicking in the speaker name and deleting, then hit Enter. Add the label and now it will only be on that one row.
Related Help Articles: