ReMASC

The ReMASC dataset is used to evaluate replay speech detection systems, containing recordings that reflect the differences between human speech and loudspeaker-based replay across various devices and acoustic environments.

Papers using ReMASC (3)

Multi-channel Replay Speech Detection using an Adaptive Learnable Beamformer2025 · 4 citesImpact of Microphone Array Mismatches to Learning-based Replay Speech Detection2025 · 2 citesMulti-Channel Replay Speech Detection using Acoustic Maps2026