在海岛湿地等环境中,声学环境较为复杂,常伴有风声、雨声、海浪声等各种噪声。为了有效解决鸟鸣声处理中的这些干扰,提高鸟类物种识别的准确性,针对海岛湿地等复杂声环境下鸟鸣声实时在线监测中的噪声干扰问题,提出了一种基于自适应卡...在海岛湿地等环境中,声学环境较为复杂,常伴有风声、雨声、海浪声等各种噪声。为了有效解决鸟鸣声处理中的这些干扰,提高鸟类物种识别的准确性,针对海岛湿地等复杂声环境下鸟鸣声实时在线监测中的噪声干扰问题,提出了一种基于自适应卡尔曼滤波‑线性预测编码(Adaptive Kalman filtering with linear predictive coding,A‑KF‑LPC)的降噪方法。通过对鸟鸣声信号进行加权滤波,增强了A‑KF‑LPC滤波的稳定性,另外采用A‑KF‑LPC滤波对噪声进行抑制,并对声信号中不确定微小片段进行精确估计,逐步逼近真实情况。通过仿真,验证了A‑KF‑LPC滤波的性能,证明其能有效降噪。实验结果表明,在不同信噪比(Signal to noise ratio,SNR)条件下,相较于传统卡尔曼滤波、最小均方误差(Least mean squares,LMS)滤波,A‑KF‑LPC滤波的鸟鸣声信号降噪方法能更有效地去除噪声;在-10 dB噪声完全覆盖信号的条件下仍能滤除部分噪声。本研究提出的A‑KF‑LPC滤波在声学信号处理领域具有重要的应用意义,为鸟类湿地生态系统研究提供了一种高效可行的解决方案,并具有潜在的应用前景。展开更多
Bird vocalizations are pivotal for ecological monitoring,providing insights into biodiversity and ecosystem health.Traditional recognition methods often neglect phase information,resulting in incomplete feature repres...Bird vocalizations are pivotal for ecological monitoring,providing insights into biodiversity and ecosystem health.Traditional recognition methods often neglect phase information,resulting in incomplete feature representation.In this paper,we introduce a novel approach to bird vocalization recognition(BVR)that integrates both amplitude and phase information,leading to enhanced species identification.We propose MHARes Net,a deep learning(DL)model that employs residual blocks and a multi-head attention mechanism to capture salient features from logarithmic power(POW),Instantaneous Frequency(IF),and Group Delay(GD)extracted from bird vocalizations.Experiments on three bird vocalization datasets demonstrate our method's superior performance,achieving accuracy rates of 94%,98.9%,and 87.1%respectively.These results indicate that our approach provides a more effective representation of bird vocalizations,outperforming existing methods.This integration of phase information in BVR is innovative and significantly advances the field of automatic bird monitoring technology,offering valuable tools for ecological research and conservation efforts.展开更多
Echolocation calls of 10 Chinese rhinolophid species were recorded to investigate the relationship between morphology and echolocation signals. All horseshoe bats use FM-CF-FM calls. Rhinolophus rex calls at 23.7 kHz,...Echolocation calls of 10 Chinese rhinolophid species were recorded to investigate the relationship between morphology and echolocation signals. All horseshoe bats use FM-CF-FM calls. Rhinolophus rex calls at 23.7 kHz, the lowest frequency in this genus. Call frequency was not correlated with body mass (P=0.200, 9 species). Close negative relationships were found between call frequency and ear length (r=-0.942, P<0.001) and also between call frequency and forearm length (r=-0.696, P<0.05). Residual analysis was carried out to remove the influence of other morphological features. After calculating ear length, forearm length residuals were not significantly related to call frequency (r=-0.095, P=0.808). The significance of the correlation between ear length and call frequency was slightly lowered (r=-0.642, P=0.062) after “removing” the influence of forearm length. Ear length, therefore, was a better predictor of call frequency than forearm length [Acta Zoologica Sinica 49(1):128-133,2003].展开更多
文摘在海岛湿地等环境中,声学环境较为复杂,常伴有风声、雨声、海浪声等各种噪声。为了有效解决鸟鸣声处理中的这些干扰,提高鸟类物种识别的准确性,针对海岛湿地等复杂声环境下鸟鸣声实时在线监测中的噪声干扰问题,提出了一种基于自适应卡尔曼滤波‑线性预测编码(Adaptive Kalman filtering with linear predictive coding,A‑KF‑LPC)的降噪方法。通过对鸟鸣声信号进行加权滤波,增强了A‑KF‑LPC滤波的稳定性,另外采用A‑KF‑LPC滤波对噪声进行抑制,并对声信号中不确定微小片段进行精确估计,逐步逼近真实情况。通过仿真,验证了A‑KF‑LPC滤波的性能,证明其能有效降噪。实验结果表明,在不同信噪比(Signal to noise ratio,SNR)条件下,相较于传统卡尔曼滤波、最小均方误差(Least mean squares,LMS)滤波,A‑KF‑LPC滤波的鸟鸣声信号降噪方法能更有效地去除噪声;在-10 dB噪声完全覆盖信号的条件下仍能滤除部分噪声。本研究提出的A‑KF‑LPC滤波在声学信号处理领域具有重要的应用意义,为鸟类湿地生态系统研究提供了一种高效可行的解决方案,并具有潜在的应用前景。
基金supported by the Beijing Natural Science Foundation (5252014)the National Natural Science Foundation of China (62303063)。
文摘Bird vocalizations are pivotal for ecological monitoring,providing insights into biodiversity and ecosystem health.Traditional recognition methods often neglect phase information,resulting in incomplete feature representation.In this paper,we introduce a novel approach to bird vocalization recognition(BVR)that integrates both amplitude and phase information,leading to enhanced species identification.We propose MHARes Net,a deep learning(DL)model that employs residual blocks and a multi-head attention mechanism to capture salient features from logarithmic power(POW),Instantaneous Frequency(IF),and Group Delay(GD)extracted from bird vocalizations.Experiments on three bird vocalization datasets demonstrate our method's superior performance,achieving accuracy rates of 94%,98.9%,and 87.1%respectively.These results indicate that our approach provides a more effective representation of bird vocalizations,outperforming existing methods.This integration of phase information in BVR is innovative and significantly advances the field of automatic bird monitoring technology,offering valuable tools for ecological research and conservation efforts.
文摘Echolocation calls of 10 Chinese rhinolophid species were recorded to investigate the relationship between morphology and echolocation signals. All horseshoe bats use FM-CF-FM calls. Rhinolophus rex calls at 23.7 kHz, the lowest frequency in this genus. Call frequency was not correlated with body mass (P=0.200, 9 species). Close negative relationships were found between call frequency and ear length (r=-0.942, P<0.001) and also between call frequency and forearm length (r=-0.696, P<0.05). Residual analysis was carried out to remove the influence of other morphological features. After calculating ear length, forearm length residuals were not significantly related to call frequency (r=-0.095, P=0.808). The significance of the correlation between ear length and call frequency was slightly lowered (r=-0.642, P=0.062) after “removing” the influence of forearm length. Ear length, therefore, was a better predictor of call frequency than forearm length [Acta Zoologica Sinica 49(1):128-133,2003].
