\u6d88\u9664\u5bf9\u8bdd\u53cc\u65b9\u6216\u591a\u65b9\u7684\u80cc\u666f\u566a\u97f3<\/strong>\uff0c\u8fd9\u8ddfANC\uff08Active Noise Cancellation\u4e3b\u52a8\u566a\u97f3\u6d88\u9664\uff09\u6d88\u9664\u8fdb\u5165\u5230\u4eba\u8033\u4e2d\u7684\u566a\u97f3\u95ee\u9898\u4e0d\u4e00\u6837\u3002<\/p>\n\n\n\n\u4f20\u7edf\u7684\u566a\u97f3\u6291\u5236\u76ee\u524d\u5df2\u7ecf\u88ab\u5e7f\u6cdb\u5e94\u7528\u4e8e\u7ec8\u7aef\u7535\u5b50\u8bbe\u5907\uff0c\u6bd4\u5982\u6211\u4eec\u6700\u5e38\u7528\u7684\u624b\u673a\u3002\u8fd9\u4e00\u6280\u672f\u5229\u7528\u5728\u4e0d\u540c\u4f4d\u7f6e\u7684\u591a\u4e2a\u9ea6\u514b\u98ce\u6765\u6355\u6349\u4eba\u58f0\uff0c\u901a\u8fc7\u8ba1\u7b97\u4e0d\u540c\u4f4d\u7f6e\u4eba\u58f0\u97f3\u91cf\u7684\u5dee\u522b\uff0c\u4ece\u800c\u5b9e\u73b0\u6e05\u6670\u7684\u8bed\u97f3\u6548\u679c\u3002\u4f46\u8fd9\u79cd\u6280\u672f\u6709\u5f88\u5927\u7684\u5c40\u9650\u6027\uff0c\u6bd4\u5982\u5f53\u4eba\u4e0d\u8bf4\u8bdd\u7684\u65f6\u5019\uff0c\u5c31\u6ca1\u529e\u6cd5\u5224\u65ad\u4eba\u58f0\u4f4d\u7f6e\uff0c\u6240\u6709\u9ea6\u514b\u98ce\u63a5\u53d7\u7684\u90fd\u662f\u73af\u5883\u566a\u97f3\uff0c\u6216\u8005\u662f\u5728\u8dd1\u6b65\u65f6\uff0c\u8ddd\u79bb\u5728\u4e0d\u505c\u53d8\u6362\uff0c\u60c5\u51b5\u5c31\u4f1a\u5f88\u590d\u6742\u3002<\/p>\n\n\n\n
\u968f\u540e\u51fa\u73b0\u4e86DSP\uff08Digital SIgnal Processing\u6570\u5b57\u4fe1\u53f7\u5904\u7406\uff09\u7b97\u6cd5\uff0c\u901a\u8fc7\u9010\u5e27\u5904\u7406\u97f3\u9891\uff0c\u5206\u8fa8\u4eba\u58f0\u548c\u566a\u97f3\uff0c\u5e76\u52a0\u4ee5\u9002\u5e94\u6765\u5b9e\u73b0\u964d\u566a\u7684\u6548\u679c\u3002\u4f46\u8fd9\u79cd\u65b9\u5f0f\u5bf9\u5904\u7406\u65e5\u5e38\u73af\u5883\u4e2d\u591a\u53d8\u7684\u566a\u97f3\u4f9d\u65e7\u6709\u5f88\u5927\u7684\u5c40\u9650\u6027\uff0c\u901a\u5e38\u6765\u8bf4\u566a\u97f3\u53ef\u4ee5\u5206\u4e3aStationary\uff08\u9759\u6b62\u7684\uff09\u548cNon-Stationary\uff08\u975e\u9759\u6b62\u7684\uff09\u4e24\u79cd\u3002<\/p>\n\n\n\n
\u4f20\u7edfDSP\u7b97\u6cd5\u53ef\u4ee5\u5f88\u597d\u5730\u5904\u7406\u9759\u6b62\u7684\u566a\u97f3<\/strong>\uff0c\u4f46\u5bf9\u4e8e\u4e00\u4e9b\u8ddf\u4eba\u58f0\u76f8\u4f3c\uff0c\u6216\u8005\u975e\u5e38\u77ed\u4fc3\u7684\u58f0\u97f3\uff08\u952e\u76d8\u6572\u51fb\u58f0\u3001\u8b66\u62a5\u58f0\u7b49\uff09\u5219\u6548\u679c\u4e0d\u5927\u3002<\/p>\n\n\n\n\u800cKrisp\u5219\u662f\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u6765\u5b9e\u73b0\u964d\u566a\uff0c\u6709\u5173\u8fd9\u65b9\u9762\u7684\u7684\u57fa\u7840\u539f\u7406\u6765\u81ea\u4e8e2015\u5e74\u7684\u300aA regression approach to speech enhancement based on deep neural networks\u300b\u8fd9\u7bc7Paper\uff0c\u8fd9\u91cc\u63d0\u5230\u4e86\u4e00\u79cdregression\u65b9\u6cd5\uff0c\u901a\u8fc7\u5b66\u4e60\u4ea7\u751f\u6bcf\u4e2a\u97f3\u9891\u7684Ratio Mask\uff0c\u505a\u51fa\u6765\u7684Ration Mask\u4fdd\u7559\u4e86\u4eba\u58f0\uff0c\u5e76\u5220\u9664\u4e86\u591a\u4f59\u7684\u566a\u97f3\u3002<\/p>\n\n\n\n
\u968f\u7740\u65f6\u95f4\u63a8\u79fb\uff0c\u8fd9\u4e2a\u7b97\u6cd5\u4e5f\u5728\u4e0d\u65ad\u5b8c\u5584\uff0c\u4f46\u4e3b\u8981\u7684\u5e95\u5c42\u903b\u8f91\u5c31\u662f\u4ee5\u4e0b\u4e09\u4e2a\u6b65\u9aa4\uff1a\u9996\u5148\u662f\u5c06\u5e72\u51c0\u7684\u8bed\u97f3\u548c\u566a\u97f3\u6df7\u5408\uff0c\u751f\u6210\u5e26\u566a\u97f3\u7684\u4eba\u58f0\u5927\u6570\u636e\u96c6\uff0c\u7136\u540e\u5c06\u8fd9\u4e9b\u6570\u636e\u96c6\u5582\u7ed9DNN\uff08Deep Neural Networks\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\uff09\u8f93\u5165\u7aef\uff0c\u7136\u540e\u5728\u8f93\u51fa\u7aef\u5219\u662f\u5582\u5bf9\u5e94\u5e72\u51c0\u7684\u4eba\u58f0\uff0c\u6700\u540e\u751f\u6210\u4e00\u4e2a\u8fc7\u6ee4\u4e86\u566a\u97f3\uff0c\u540c\u65f6\u4fdd\u7559\u4e86\u4eba\u58f0\u7684Mask\u3002<\/p>\n\n\n\n