MDCNN-SID: Multi-scale Dilated Convolution Network for Singer Identification
This addresses singer identification for music analysis applications, but it is incremental as it builds on existing methods by shifting to the waveform domain.
The paper tackled singer identification by proposing an end-to-end architecture in the waveform domain to avoid information loss from spectral transformations, achieving comparable performance on the Artist20 benchmark dataset with significant improvements over related works.
Most singer identification methods are processed in the frequency domain, which potentially leads to information loss during the spectral transformation. In this paper, instead of the frequency domain, we propose an end-to-end architecture that addresses this problem in the waveform domain. An encoder based on Multi-scale Dilated Convolution Neural Networks (MDCNN) was introduced to generate wave embedding from the raw audio signal. Specifically, dilated convolution layers are used in the proposed method to enlarge the receptive field, aiming to extract song-level features. Furthermore, skip connection in the backbone network integrates the multi-resolution acoustic features learned by the stack of convolution layers. Then, the obtained wave embedding is passed into the following networks for singer identification. In experiments, the proposed method achieves comparable performance on the benchmark dataset of Artist20, which significantly improves related works.