An Improved Approach to Audio Segmentation and Classification in Broadcasting Industries

Audio segmentation and classification are the basis of audio processing in broadcasting industries. A Dual-CNN (Dual-Convolutional Neural Network) method is proposed in this article in which it is possible to pre-train a CNN with unlabeled audio data so as to deal with the scarcity of labeled data. Auto-encoders (including an encoder and a decoder) are utilized, thus the name “Dual.” In the first place, audio sampling points and the derived STFT (Short-Time Fourier Transform) spectrograms pass through their own CNNs. Fusion of the extracted features is then performed. Finally, the merged features are sent to a fully connected network and the classification results are produced via Softmax. Being one of the segmentation-by-classification approaches, our solution also presents a novel smoothing method (SEG-smoothing) in order to deliver the best result of segmentation. A series of experiments have been conducted and their result verifies that the proposed approach for segmentation and classification outperforms alternative solutions.

Tools for Automatic Audio Management

The purpose of this chapter is to introduce tools for automatic audio management. The authors present applications which are already available for the users and describe the algorithms and methods behind these applications and their performance. They also discuss the concept of metadata, which is an important prerequisite for modern distributed personal content applications. The variety of automatic audio management tools is wide-ranging. This chapter covers audio segmentation and classification, query by example of audio, music retrieval and recommendation, and speech management, which they consider as being the most important aspects of audio information management. Computational complexity is one major concern in the present era of personal mobile devices and large multimedia collections available on the internet. Therefore they also introduce clustering and indexing techniques which are developed for faster access in large databases.