Novel Audio Feature Projection Using KDLPCCA-Based Correlation with EEG Features for Favorite Music Classification

Abstract

A novel audio feature projection using Kernel Discriminative Locality Preserving Canonical Correlation Analysis (KDLPCCA)-based correlation with electroencephalogram (EEG) features for favorite music classification is presented in this paper. The projected audio features reflect individual music preference adaptively since they are calculated by considering correlations with the user's EEG signals during listening to musical pieces that the user likes/dislikes via a novel CCA proposed in this paper. The novel CCA, called KDLPCCA, can consider not only a non-linear correlation but also local properties and discriminative information of each class sample, namely, music likes/dislikes. Specifically, local properties reflect intrinsic data structures of the original audio features, and discriminative information enhances the power of the final classification. Hence, the projected audio features have an optimal correlation with individual music preference reflected in the user's EEG signals, adaptively. If the KDLPCCA-based projection that can transform original audio features into novel audio features is calculated once, our method can extract projected audio features from a new musical piece without newly observing individual EEG signals. Our method therefore has a high level of practicability. Consequently, effective classification of user's favorite musical pieces via a Support Vector Machine (SVM) classifier using the new projected audio features becomes feasible. Experimental results show that our method for favorite music classification using projected audio features via the novel CCA outperforms methods using original audio features, EEG features and even audio features projected by other state-of-the-art CCAs.

Human-Centered Favorite Music Classification Using EEG-Based Individual Music Preference Via Deep Time-Series CCA

“…4 and Table 3. Specifically, we compared the following seven methods explained in Table 3 [17] which used KDLPCCA and the classifier based on the Support Vector Machine (SVM) [23]. Note that the architectures of MLPs used for CCA and classifier were the networks without recurrent connections.…”

“…We used this evaluation results as labels for EEG feature selection. The rest conditions of our experiments, e.g., used EEG channels, equipment and so on, are described in our previous paper [17].…”

“…Namely, classification of the user's favorite musical pieces becomes effectively feasible by using DTCCA as feature extractor and DNN as classifier. [17]. Since the above feature selection algorithm [16] utilizes class information represnting favorite or unfavorite music of the user, our method try to handle the user's music preference via this selected EEG features.…”

“…The kinds of EEG and audio feature type are summarized in Table 1 and 2. Note that the rest details of calculating these features and discussion for those are given in our previous paper [17] to save the space of this paper.…”

Supervised Fractional-Order Embedding Geometrical Multi-View CCA (SFGMCCA) for Multiple Feature Integration

“…Furthermore, for considering the nonlinear structure directly, kernel CCA [14] was proposed, and Sun et al derived kernelized version of DCCA (KDCCA) [7]. Moreover, kernelized version of DLPCCA (KDLPCCA) [15] which can consider the non-linear structure, locality preserving and discriminant analysis was proposed.…”