Example of simulated bat-call recording from known trajectory

The archive "ExampleOfSimulatedRecordingFromKnownTrajectory.ZIP" contains an example of virtual bat Myotis myotis, flying along some random trajectory and emitting an echolocation sequence, recorded by 4 microphones, placed near this trajectory. This example is to compare our method of reconstruction of trajectory with a baseline. The trajectory is chosen to contain 1 point where echo is stronger than a direct signal (point 3), 1 point baseline fails to reconstruct (point 5) and other points where calculation of TDOAs by baseline is good (close to real TDOA). Subdirectory "whole_simulated_signal" contains signal as it is emitted by bat and as it is received by all 4 microphones, recording both the direct signal and the echo. Subdirectory "parts_of_signal,recorded_during_fly_near_each_point" contains division of echolocation sequence into 5 parts, emitted by bat while flying near each of 5 trajectory points. Direct signal and Echo are plotted in black and red correspondingly. At the end of signal (point 5) there is an additional noise, to make reconstruction of this point difficult. Supplementary subdirectory "separately_echo_and_direct_signal" contains zoom on smaller parts of a signal (time-segments, used for calculation of TDOAs and FDOAs). Echoes and Direct signal are plotted in 2 different colors. Subdirectories "pointX_time-segments_and_resulting_TFDOAs" reflect the process of calculation of TDOAs and FDOAs as correlations of signals of different microphones on the same time-segment. Some of images correspond to images in the paper "Joint time delay and Doppler passive acoustic 3D tracking, applied to studies of bats in natural habitats." by Ales Mishchenko and Herve Glotin Subdirectory "trajectory" contains resulting TDOAs in the form of distance difference, as well as the resulting trajectory The archive "ExampleOfSimulatedRecordingFromKnownTrajectory.ZIP" duplicates information in images(.PNG) and figures in Matlab format (.FIG), as well as in tables(.CSV) for small data or in Matlab data files(.MAT) for big data ------------- For additional information contact Ales Mishchenko ([e-mail])压缩包"ExampleOfSimulatedRecordingFromKnownTrajectory.ZIP"包含一份仿真示例数据：针对小鼠耳蝠（Myotis myotis）的虚拟场景，该蝙蝠沿随机轨迹飞行并发射回声定位序列，由4台布设于轨迹附近的麦克风录制得到。本示例用于对比我们提出的轨迹重构方法与基线方法的性能。该轨迹设置了1个回声强于直达信号的点位（点位3）、1个基线方法无法完成重构的点位（点位5），以及其余若干基线方法可准确计算到达时间差（TDOA，Time Difference of Arrival）且结果接近真实值的点位。 子目录"whole_simulated_signal"中存储了蝙蝠发射的原始信号，以及4台麦克风接收到的包含直达信号与回声的全量录制信号。 子目录"parts_of_signal,recorded_during_fly_near_each_point"将回声定位序列划分为5个片段，对应蝙蝠在5个轨迹点位附近飞行时发射的信号。其中直达信号与回声分别以黑色与红色绘制。在信号末端（点位5）添加了额外噪声，以增加该点位的重构难度。 补充子目录"separately_echo_and_direct_signal"中展示了用于计算到达时间差（TDOA）与多普勒频率差（FDOA，Frequency Difference of Arrival）的信号局部放大时域片段。回声与直达信号以两种不同颜色区分绘制。 子目录"pointX_time-segments_and_resulting_TFDOAs"展示了针对同一时域片段的不同麦克风信号进行互相关计算，以得到时频差（TFDOAs）的过程。其中部分图像对应Ales Mishchenko与Herve Glotin发表的论文《Joint time delay and Doppler passive acoustic 3D tracking, applied to studies of bats in natural habitats.》中的配图。 子目录"trajectory"中存储了以距离差形式呈现的到达时间差（TDOA）计算结果，以及最终重构出的轨迹。 本压缩包"ExampleOfSimulatedRecordingFromKnownTrajectory.ZIP"同时提供该数据集的多种格式素材：PNG图像文件、MATLAB格式图形文件（.FIG）、小型数据集对应的逗号分隔值表格文件（.CSV），以及大型数据集对应的MATLAB数据文件（.MAT）。 —— 如需获取更多信息，请联系Ales Mishchenko（邮箱：[e-mail]）