HyperAI초신경

Speech Enhancement On Ears Wham

평가 지표

DNSMOS
ESTOI
PESQ-WB
POLQA
SI-SDR
SIGMOS

평가 결과

이 벤치마크에서 각 모델의 성능 결과

비교 표
모델 이름DNSMOSESTOIPESQ-WBPOLQASI-SDRSIGMOS
hybrid-transformers-for-music-source3.660.712.372.9716.922.87
tasnet-surpassing-ideal-time-frequency3.470.702.312.7316.932.69
speech-enhancement-and-dereverberation-with3.880.732.503.4016.783.41
investigating-training-objectives-for3.720.733.093.7116.293.18
schrodinger-bridge-for-generative-speech3.830.732.333.4617.853.44
conditional-diffusion-probabilistic-model-for2.870.531.601.818.352.08