Librispeech-960

Librispeech-960 is a large-scale dataset used for evaluating automatic speech recognition systems, containing 960 hours of English speech data derived from audiobooks.

Papers using Librispeech-960 (10)

A Comparative Analysis on ASR System Combination for Attention, CTC, Factored Hybrid, and Transducer Models2025IML-Spikeformer: Input-aware Multi-Level Spiking Transformer for Speech Processing2025Right Label Context in End-to-End Training of Time-Synchronous ASR Models2025Make More of Your Data: Minimal Effort Data Augmentation for Automatic Speech Recognition and Translation2022 · 1 citesContrastive Siamese Network for Semi-supervised Speech Recognition2022BayesSpeech: A Bayesian Transformer Network for Automatic Speech Recognition2023Competitive and Resource Efficient Factored Hybrid HMM Systems are Simpler Than You Think2023Fast-HuBERT: An Efficient Training Framework for Self-Supervised Speech Representation Learning2023Contextualized Automatic Speech Recognition with Attention-Based Bias Phrase Boosted Beam Search2024Investigating the Effect of Label Topology and Training Criterion on ASR Performance and Alignment Quality2024