MMedFD: A Real-World Healthcare Benchmark for Multi-Turn Full-Duplex ASR

MMedFD is a benchmark dataset for multi-turn, full-duplex automatic speech recognition in real-world healthcare settings. The dataset was created by HanselZz and was last updated on Hugging Face in February 2026. Full access requires internal approval and a research-only data use agreement.

Use Cases

• Training full-duplex ASR models based on multi-turn conversational audio described in the benchmark.
• Evaluating speech recognition robustness based on real-world healthcare environment recordings.
• Developing dialogue systems for clinical settings based on the multi-turn interaction structure.

Strengths

• Designed as a real-world benchmark, suggesting practical relevance.
• Focuses on multi-turn and full-duplex interactions, which are complex ASR scenarios.
• Last updated on 2026-02-25, indicating recent maintenance.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and sample data are unknown, which may limit suitability assessment.

Provenance

Source

HanselZz on Hugging Face.

Collection Method

Likely collected from real-world healthcare interactions, but specific gathering method is not detailed.

Time Range

null

Freshness

Last updated 2026-02-25 05:24:22; freshness should be verified.

Geography

null