"Wow!" Bayesian surprise for salient acoustic event detection

We propose the use of Bayesian surprise to detect arbitrary, salient acoustic events. We use Gaussian or Gamma distributions to model the spectrogram distribution and use the Kullback-Leibler divergence of the posterior and prior distribution to calculate how “unexpected” and thus surprising newly observed audio samples are. This way, we efficiently detect arbitrary surprising/salient acoustic events. MOTIVATION • identify subsets within sensory inputs that are likely to contain important information • focus complex processing operations on the selected, potentially relevant information • in general, drastically reduce the computational requirements to process data • real-time processing and reflex-like reactions despite computational restrictions PRINCIPLE • an observed spectrogram element G(t, ω) is “surprising” if the updated (using Bayes’ rule) distribution P post differs significantly from the prior distribution P prior SA(t, ω) = DKL(P ω post||P prior) (1) = ∫ P post log Pω post Pω prior dg (2) with Kullback-Leiber divergence DKL • surprise at time t over all frequencies