Mandarin

The 'Mandarin' dataset/benchmark is used to evaluate the effectiveness of speech representation learning methods, particularly in the context of tonal languages, by assessing their ability to handle speaker variation and tone distinctions.

Papers using Mandarin (16)

Language Barriers: Evaluating Cross-Lingual Performance of CNN and Transformer Architectures for Speech Quality Estimation2025 · 1 citesToward Unified Chinese Multi-Dialectal Speech Recognition via Pinyin Intermediate Representation2026Prosodic ABX: A Language-Agnostic Method for Measuring Prosodic Contrast in Speech Representations2026Rethinking Entropy Allocation in LLM-based ASR: Understanding the Dynamics between Speech Encoders and LLMs2026SITA: Learning Speaker-Invariant and Tone-Aware Speech Representations for Low-Resource Tonal Languages2026Unsupervised lexicon learning from speech is limited by representations rather than clustering2025HENT-SRT: Hierarchical Efficient Neural Transducer with Self-Distillation for Joint Speech Recognition and Translation2025A Self-Refining Framework for Enhancing ASR Using TTS-Synthesized Data2025Analyzing Acoustic Word Embeddings from Pre-trained Self-supervised Speech Models2022 · 1 citesMulti-pass Training and Cross-information Fusion for Low-resource End-to-end Accented Speech Recognition2023Effects of Convolutional Autoencoder Bottleneck Width on StarGAN-based Singing Technique Conversion2023Accent-VITS:accent transfer for end-to-end TTS2023Period Singer: Integrating Periodic and Aperiodic Variational Autoencoders for Natural-Sounding End-to-End Singing Voice Synthesis2024PRESENT: Zero-Shot Text-to-Prosody Control2024LA-RAG:Enhancing LLM-based ASR Accuracy with Retrieval-Augmented Generation2024Do Discrete Self-Supervised Representations of Speech Capture Tone Distinctions?2024