使用RNN模型作語音訊號的除噪 Original Voice Clip:錄下的人聲語音訊號 Office Noise:錄下的辦公室背景噪音 Mixing:前2項混合而成的訊號 After RNN:透過RNN模型除噪後的語音訊號波形 By Directional Mic in F15K:筆電上指向型麥克風的所錄下除噪後的波形 結論:透過RNN模型除噪的功能,具有近似指向型麥克風的除噪功能。也就是以軟體處理的技術來取代實體裝置。 From lower right 2 pictures, voice clip mixed with background noise was restored well in comparison to audio waveform recording by directional mic and original voice clip waveform. Ref: https://people.xiph.org/~jm/demo/rnnoise/ https://hacks.mozilla.org/2017/09/rnnoise-deep-learning-noise-suppression/ https://github.com/xiph/rnnoise mic denoise rnn audacity: import -> raw data -> Signed 16bit/Little endian/one channel, 48000 rnn frank@frank-GL753VD:~/1T/back0529/mic import format: Signed 16 bit mono format frank@frank-GL753VD:~/1T/back0529/mic ...