Domestic Cat Sound Classification Using Learned Features from Deep Neural Nets

Abstract

The domestic cat (Feliscatus) is one of the most attractive pets in the world, and it generates mysterious kinds of sound according to its mood and situation. In this paper, we deal with the automatic classification of cat sounds using machine learning. Machine learning approach for the classification requires class labeled data, so our work starts with building a small dataset named CatSound across 10 categories. Along with the original dataset, we increase the amount of data with various audio data augmentation methods to help our classification task. In this study, we use two types of learned features from deep neural networks; one from a pre-trained convolutional neural net (CNN) on music data by transfer learning and the other from unsupervised convolutional deep belief network that is (CDBN) solely trained on a collected set of cat sounds. In addition to conventional GAP, we propose an effective pooling method called FDAP to explore a number of meaningful features. In FDAP, the frequency dimension is roughly divided and then the average pooling is applied in each division. For the classification, we exploited five different machine learning algorithms and an ensemble of them. We compare the classification performances with respect following factors: the amount of data increased by augmentation, the learned features from pre-trained CNN or unsupervised CDBN, conventional GAP or FDAP, and the machine learning algorithms used for the classification. As expected, the proposed FDAP features with larger amount of data increased by augmentation combined with the ensemble approach have produced the best accuracy. Moreover, both learned features from pre-trained CNN and unsupervised CDBN produce good results in the experiment. Therefore, with the combination of all those positive factors, we obtained the best result of 91.13% in accuracy, 0.91 in f1-score, and 0.995 in area under the curve (AUC) score.

Data augmentation approaches for improving animal audio classification

“… BIRDZ, the control and real-world audio dataset used in [24]. The real-world recordings were  CAT, the cat sound dataset was presented in [28,50]. In the following Table 1 we report the performance obtained by the four data augmentation protocols, comparing them with no augmentation (NoAUG) as baseline.…”

Adaptive Multi-Scale Detection of Acoustic Events

“…Besides, multi-scale feature extraction in frequency level is also important [56]. Intuitively, different frequency ranges dominate different acoustic events, which makes it hard to predict accurate results only with certain frequency range.…”

Sound Source Distance Estimation Using Deep Learning: An Image Classification Approach

“…The goal of this study is to use a deep learning model to learn relevant distance-related features from the audio signals for classification. To accomplish other deep learning-based audio analysis goals, previous researchers [36,41] have represented the time domain audio signals by mel [33] spectrograms before training. The mel spectrogram is a time-frequency representation of an audio signal that compresses high-frequency components and focuses more on low-frequency components [36].…”