本文探讨了独立成分分析(ICA)的基本原理和方法，并通过Matlab处理音频记录。实验中，使用不同位置的麦克风或线性组合的方式录制混合音频信号，然后应用ICA进行信号恢复。实验结果显示，虽然快速ICA可以实现声音分离，但效果并不十分理想。尝试添加噪声或不同类型的音频源（如音乐、敲击声）时，分离效果有所改善。
Independent Component Analysis (ICA)
Zhang Bing 171848
Purpose：
1. Study the concept and methods of ICA in the handout and the text book
2. Explore the audio recording the processing functions of Matlab
3. Record mixed signals using one of the following methods:
a) Select at least 3 audio sources and use the equivalent number of microphones at different positions to record the multiple channels of mixed audio signals for 5 ~ 10 seconds simultaneously.
b) Select at least 3 audio sources and use one microphone to record the audio signal from each source for 5 ~ 10 seconds, then generate the equivalent number of mixed signals with different linear combinations of the recorded signals
Note: method a) is preferred, because it is more “real”.
4. Apply ICA on the mixed audio signals and see whether the original
signals can be recovered successfully
successfully
最低0.47元/天 解锁文章
扫一扫
4万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
