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 4 救护车噪声5种信噪比下各模型的语音增强效果对比

Tab.4 Comparison of speech enhancement effect of various model under five signal-to-noise ratios of ambulance 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 1.76 2.09 2.28 2.37 2.55 −10 −5 0 5 10 AV-ConvTasnet 2.27 2.33 2.42 2.47 2.52 3.97 3.96 4.20 4.04 4.06 MuSE 2.32 2.42 2.54 2.62 2.67 5.92 6.13 6.47 6.24 6.17 AV-Sepformer 2.69 2.74 2.84 2.90 2.93 6.75 6.58 6.89 6.77 6.81 本文模型 2.96 3.10 3.28 3.39 3.47 13.93 14.07 14.80 14.58 14.68