The speaker labels during the meeting isn't accurate
During the meeting, you may find that the speaker labels are inaccurate, or you may find that someone was clearly speaking, but Caption.Ed Scribe labels the paragraph as "Unknown speakers."
During the meeting, Caption.Ed Scribe will try to identify the speakers based on the audio inputs it detects. However, once the session has finished, additional processing happens to identify the correct speaker and assign that speaker to the correct paragraph.
Therefore, you will find that the speaker labels are correctly identified in the transcript after your session.