Research Article

Multichannel Speech Enhancement in Vehicle Environment Based on Interchannel Attention Mechanism

Table 1

Evaluation results of our proposed model compared with other methods on the same dataset. Four metrics and two SNRs are considered.

Method# MicsSDRSI-SNRPESQSTOI
−10 dB−5 dB−10 dB−5 dB−10 dB−5 dB−10 dB−5 dB

Noise2 linear−6.720.34−7.190.241.071.130.400.64
4 linear−5.930.11−6.400.021.061.110.400.64
2 × 2 dB−6.720.32−6.180.221.071.130.400.64
4 dB−6.890.68−7.170.601.071.130.460.70
+NCC2 linear7.4511.986.4211.301.201.760.730.87
4 linear8.8912.447.8611.751.281.830.750.88
2 × 2 dB7.6313.147.5012.371.231.850.710.89
4 dB8.3814.587.6013.961.302.020.760.91
+ICD2 linear7.4211.956.3911.271.191.750.720.86
4 linear8.8012.387.7911.681.271.820.740.87
2 × 2 dB7.5513.067.4412.301.221.830.710.89
4 dB8.3314.517.5613.931.272.010.760.91
+MCS2 linear7.4011.926.3611.251.191.750.720.86
4 linear8.8312.407.8311.721.271.820.740.87
2 × 2 dB7.5213.047.4212.281.211.820.700.88
4 dB8.2814.467.5213.791.252.000.750.90
Proposed2 linear7.5212.066.4811.361.221.770.740.88
4 linear8.9412.497.9111.831.291.830.750.88
2 × 2 dB7.7213.237.5812.451.241.860.720.90
4 dB8.4114.667.6614.071.332.040.760.91

The best evaluation results are shown in bold, comparing the results of the four speech enhancement methods used in the four microphone arrays.