Improving Deep Learning Sound Events Classifiers using Gram Matrix Feature-wise Correlations
This is an incremental improvement for sound event classification tasks.
The authors tackled sound event classification by analyzing CNN activations with Gram matrices to compute similarity metrics across classes, and their method improved baseline models across four architectures on two datasets.
In this paper, we propose a new Sound Event Classification (SEC) method which is inspired in recent works for out-of-distribution detection. In our method, we analyse all the activations of a generic CNN in order to produce feature representations using Gram Matrices. The similarity metrics are evaluated considering all possible classes, and the final prediction is defined as the class that minimizes the deviation with respect to the features seeing during training. The proposed approach can be applied to any CNN and our experimental evaluation of four different architectures on two datasets demonstrated that our method consistently improves the baseline models.