What is diarisation?

Diarisation is the process of automatically identifying the speakers in a given audio or video recording. It enables the creation of a transcription that differentiates between the speakers, making it easier for accurate labeling and indexing of the information in the spoken data. Diarisation is essential in various spoken language processing applications such as speech recognition, speaker identification, and speech segmentation. The process can be achieved using different approaches, including acoustic-based, lexical-based, and speaker-model-based methods. The accuracy of diarisation relies on various factors such as the quality of the audio recording, the number of speakers, and background noise in the audio recording.