Japanese

The 'Japanese' dataset/benchmark contains high-quality human-labeled, pseudo-labeled, and synthetic data used to evaluate the performance of automatic speech recognition models specialized for the Japanese language.

Papers using Japanese (4)

LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context2025 · 3 citesCC-G2PnP: Streaming Grapheme-to-Phoneme and prosody with Conformer-CTC for unsegmented languages2026Flavors of Moonshine: Tiny Specialized ASR Models for Edge Devices2025Contextualized Automatic Speech Recognition with Dynamic Vocabulary2024