11 Spectral Analysis Part 1 - Intro to Neural Computation
MICHALE FEE: So who remembers what that's called? A spectrograph. Good. And it's a spectrogram of me. Good morning, class. OK. So it's a spectrogram of speech. And so we are going to continue today on the topic of understanding-- developing methods for understanding how to characterize and understand temporally structured signals. So that is the microphone recording of my voice saying "good morning, class." And then this is a spectrogram of that signal where at each moment in time, you can actually extract the spectral structure of that signal. And you can see that the information in speech signals is actually carried in parts of the signal in the way the power in the signal at different frequencies changes over time. And your ears detect these changes in frequency and translate that into information about what I'm saying. And so we're going to today start on a-- well, we sort of started last time, but we're really going to get going on t...