This page contains audio samples for the 35 components learned by LISE from pretrained speaker embeddings. Components are ordered by perceptual discriminability.