Fig. 11
From: Unsupervised adaptation of PLDA models for broadcast diarization
Semi-supervised adaptation strategy based on the longitudinal unsupervised adaptation approach for the episodes 1 and 2. \(\mathcal {M}_{\text {OOD}}\) represents the out-of-domain model, \(\mathcal {M}_{\text {ID}}\) depicts the show in-domain model and \(\mathcal {M}_{1}\) and \(\mathcal {M}_{2}\) the adapted models for each episode. ΦID and ΘID symbolize the in-domain i-vectors and its speaker labels. Φ1 and Φ2 are the i-vectors from each episode