Multi Feature Sound Classification using Deep Learning

Artificial Intelligence plays an important role in acoustics recognition. Importantly, being able to automatically and accurately identify environmental sounds opens up a broad range of applications. Deep learning techniques can assist in the recognition of sounds which we come across in our day-to-day life. Most of the previous work in environmental sound classification involves training a model on a single set of features. Convolutional neural network (ConvNet) is a class of deep feed-forward neural network which exploits the strong spatially local correlation in natural images. It achieves successful performance in visual analyzing area. This paper primarily focuses on two key aims: the first aim is to perform a multilabel classification system and the second aim is to develop Stacked Bidirectional Long Short-Term Memory (LSTM) with two hidden layers to categorize multiple UAVs sounds. There are three portions to perform environmental classification. Firstly, the input signal is converted into spectrogram image with time-frequency representation using short time Fourier transforms. Secondly, this spectrogram is used to extract features with local binary pattern of three different radius and neighborhood sizes. The three distinct features resulted from local binary pattern based on spectrogram are concatenated and used as one feature vector.