计算机工程与应用
計算機工程與應用
계산궤공정여응용
Computer Engineering and Applications
2015年
20期
203-207,212
,共6页
自适应%语音端点检测%Mel能量%时频参数
自適應%語音耑點檢測%Mel能量%時頻參數
자괄응%어음단점검측%Mel능량%시빈삼수
self-adaptive%voice activity detection%Mel-scale log-energy%Time-Frequency(TF)parameter
为了解决低信噪比环境下传统的语音端点检测算法性能较差且不能自适应环境噪声，提出了一种基于时频参数融合的自适应语音端点检测算法。将对数能量与改进的Mel能量进行融合，获得了一种新的时频参数（TF），该参数能有效地区分语音段和噪声段。使用该参数在噪声段对阈值进行更新，采用门限检测法判定出语音端点。仿真实验表明，该算法具有较好的鲁棒性，且能够准确地检测出语音端点。当信噪比（SNR）为0 dB时，端点检测错误率仅为15%左右。
為瞭解決低信譟比環境下傳統的語音耑點檢測算法性能較差且不能自適應環境譟聲，提齣瞭一種基于時頻參數融閤的自適應語音耑點檢測算法。將對數能量與改進的Mel能量進行融閤，穫得瞭一種新的時頻參數（TF），該參數能有效地區分語音段和譟聲段。使用該參數在譟聲段對閾值進行更新，採用門限檢測法判定齣語音耑點。倣真實驗錶明，該算法具有較好的魯棒性，且能夠準確地檢測齣語音耑點。噹信譟比（SNR）為0 dB時，耑點檢測錯誤率僅為15%左右。
위료해결저신조비배경하전통적어음단점검측산법성능교차차불능자괄응배경조성，제출료일충기우시빈삼수융합적자괄응어음단점검측산법。장대수능량여개진적Mel능량진행융합，획득료일충신적시빈삼수（TF），해삼수능유효지구분어음단화조성단。사용해삼수재조성단대역치진행경신，채용문한검측법판정출어음단점。방진실험표명，해산법구유교호적로봉성，차능구준학지검측출어음단점。당신조비（SNR）위0 dB시，단점검측착오솔부위15%좌우。
In order to solve the inferior performance and sad self-adaptive of the traditional voice activity detection algo-rithm in an environment with low Signal to Noise Ratio(SNR), a new self-adaptive voice activity detection algorithm based on TF parameters is put forward. After introducing the time-domain log-energy and improved mel-scale energy, the new Time-Frequency(TF)parameters are acquired by coalescing them, which make it possible for distinguishing speech from noise effectively. Then, the TF parameters are updated to predicate endpoint through the threshold test. Simulation experiments show that the algorithm has better robustness and more precise detection. When the SNR is 0 dB, the error rate of the algorithm is about 15%.
