FMSU
Collection
Towards Fine-Grained Multi-Dimensional Speech Understanding: Data Pipeline, Benchmark, and Model • 2 items • Updated
The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.
A pioneering Fine-grained Multi-dimensional Speech Understanding Benchmark. Comprising over 20,000 bilingual instances, FMSU-Bench systematically covers 14 distinct speech dimensions structured into a comprehensive 5-tier taxonomy: Speaker Demographics, Acoustic-Prosodic Features, Affective and Semantic Reasoning, Acoustic Scene Analysis, Linguistic-Paralinguistic Integration.
- Downloads last month
- -