Is there any method to get rid of noise in wite circles of the spectrogram? I have done the high pass filter for a cut-off frequency of 1 kHz. And I'v also done a adptive noise reduction in adbobe audition. But there are still some noise between bird-songs.
Can I get rid of them and get a cleaned spectrogram, which only have target object. Any thoughts on this problem?
Without much context about your audio signals, this looks like uniformly distributed low energy noise. Linear filtering may not work very well here, but a simple non-linear method can be suggested. Basically define a threshold of meaningful signal energy within a frequency band and then ignore frequency contributions below that threshold. Something like:
# Frequency analysis for non-linear filtering
X_w = np.fft.fft(x_t)
# Filter low energy noise components
max_freq_comp = np.max(np.abs(X_w))
X_w_ignore_low = np.where(np.abs(X_w) > max_freq_comp/10, X_w, 0+0j)
# Filter DC component
X_w_ignore_low[0] = 0+0j
You can check the full python program listing, I responded to a similar question here
