A deliberately tiny, instant-loading reverberant speech-like test set — the quick sanity check for dereverberation and robust-ASR front-ends. Every item is a (dry, wet) pair: a clean synthetic speech-like source and its reverberant version through a real-synthetic room impulse response (RIR), with the room's ground-truth acoustics attached. sanity run is cheap. Stratification (balanced across T60 × DRR bins): - t60bin: T60low (<0.3 s) 155 · T60mid (0.3–0.8 s) 143 · T60high (0.8–1.5 s) 93 · T60vhigh (≥1.5 s) 109 - drrbin: DRRvlow (<−8 dB) 163 · DRRlow (−8…−2) 157 · DRRmid (−2…6) 91 · DRRhigh (≥6) 89 Dry sources are synthetic speech-LIKE probe signals, not real speech: seeded harmonic…
Publicly accessible
cc-by-4.0
n<1K
A small, standardized audio-robustness evaluation set packaged to drop into eval loops. It pairs fully synthetic, license-clean speech-like probe signals with 300 real-synthetic room impulse responses (RIRs) and a labeled degradation chain (reverberation + additive noise), so you can measure how a model's acoustic predictions hold up as rooms get more reverberant and noisier. Think of it as a quick sanity/robustness check: one line to load, a concrete metric, and ground-truth room acoustics for every clip. Measured ground-truth ranges (from the underlying RIRs): - reverbbin distribution: mild 875, strong 600, clean 25 Sources are synthetic speech-LIKE probe signals, not real speech. Each is…
Publicly accessible
cc-by-4.0
1K<n<10K
