Detecting Cover Songs with Pitch Class Key-Invariant Networks

Abstract

Deep Learning (DL) has recently been applied successfully to the task of Cover Song Identification (CSI). Meanwhile, neural networks that consider music signal data structure in their design have been developed. In this paper, we propose a Pitch Class Key-Invariant Network, PiCKINet, for CSI. Like some other CSI networks, PiCKINet inputs a Constant-Q Transform (CQT) pitch feature. Unlike other such networks, large multi-octave kernels produce a latent representation with pitch class dimensions that are maintained throughout PiCKINet by key-invariant convolutions. PiCK-INet is seen to be more effective, and efficient, than other CQT-based networks. We also propose an extended variant, PiCKINet+, that employs a centre loss penalty, squeeze and excite units, and octave swapping data augmentation. PiCK-INet+ shows an improvement of ∼17% MAP relative to the well-known CQTNet when tested on a set of ∼16K tracks.

DisCover: Disentangled Music Representation Learning for Cover Song Identification

“…• CQTNet [63]: CQTNet uses carefully designed kernels and dilated convolutions to extend the receptive field, which can improve the model's representation learning capacity. • PICKiNet [33]: PICKiNet devises pitch class blocks to obtain the key-invariant musical features.…”

Analysis of Social Media Messaging for Health Promotion by the CDC from 2018 to2019 in Guangzhou, China (Preprint)

“…Byte-Cover [6] and ByteCover2 [7] achieved state-of-the-art results with a ResNet-IBN50 [8] backbone and multi-loss training with cross-entropy and triplet loss. PiCKINet [9] proposed Pitch Class Blocks in order to maintain the key-invariance features of music. LyraC-Net [10] utilized WideResNet as the backbone, and combined classification and metric learning for optimization.…”