TED-LIUMv-3

TED-LIUMv3 is a dataset that contains transcriptions of TED Talks and is used to evaluate the effectiveness of automatic context discovery strategies in improving transcription accuracy for Automatic Speech Recognition (ASR) systems.

Papers using TED-LIUMv-3 (3)

SegAug: CTC-Aligned Segmented Augmentation For Robust RNN-Transducer Based Speech Recognition2025 · 2 citesRetrieval Augmented Generation based context discovery for ASR2025 · 1 citesTwo-Stage Adaptation for Non-Normative Speech Recognition: Revisiting Speaker-Independent Initialization for Personalization2026