Soundscape researchers use spectrograms to visually identify species and scan through hours of audio data. Using machine learning and neural networks, they pick out specific audio signals from ...