基于卷积和门控注意的两阶段视听语音增强算法

王盼蓉,贾海蓉,段淑斐

Two-stage audio-visual speech enhancement algorithm based on convolution and gated attention

Panrong WANG,Hairong JIA,Shufei DUAN

表 3 风雨噪声5种信噪比下各模型的语音增强效果对比

Tab.3 Comparison of speech enhancement effect of various model under five signal-to-noise ratios of wind and rain noise

模型 PESQ SNR/dB −10 dB −5 dB 0 dB 5 dB 10 dB −10 dB −5 dB 0 dB 5 dB 10 dB Noisy 1.54 1.67 1.86 2.15 2.47 −10 −5 0 5 10 AV-ConvTasnet 2.13 2.17 2.32 2.36 2.40 3.98 4.26 4.44 4.42 4.31 MuSE 2.19 2.25 2.37 2.42 2.46 4.57 5.76 6.35 6.41 6.29 AV-Sepformer 2.53 2.59 2.75 2.82 2.88 6.47 6.58 6.83 6.89 6.78 本文模型 2.77 2.79 3.01 3.16 3.25 12.29 13.09 14.11 14.54 14.55