DEMAND

The 'DEMAND' dataset is a benchmark that contains recordings of speech in various noise conditions and is used to evaluate the performance of speech enhancement algorithms.

Papers using DEMAND (3)

Perceptual Loss with Recognition Model for Single-Channel Enhancement and Robust ASR2021 · 8 citesGLD-Net: Improving Monaural Speech Enhancement by Learning Global and Local Dependency Features with GLD Block2022Array Geometry-Robust Attention-Based Neural Beamformer for Moving Speakers2024