Detecting musical versions (different renditions of the same piece) is a challenging task with important applications. Because of the ground truth nature, existing approaches match musical versions at the track level (e.g., whole song). However, most applications require to match them at the segment level (e.g., 20s chunks). In addition, existing approaches resort to classification and triplet losses, disregarding more recent losses that could bring meaningful improvements. In this paper, we propose a method to learn from weakly annotated segments, together with a contrastive loss variant that outperforms well-studied alternatives. The former is based on pairwise segment distance reductions, while the latter modifies an existing loss following decoupling, hyper-parameter, and geometric considerations. With these two elements, we do not only achieve state-of-the-art results in the standard track-level evaluation, but we also obtain a breakthrough performance in a segment-level evaluation. We believe that, due to the generality of the challenges addressed here, the proposed methods may find utility in domains beyond audio or musical version matching.

Navigating large audio collections presents a significant challenge due to the intricate nature of sound properties and the varied needs of users. To enhance user experience, audio-sharing platforms offer a sound similarity function, which leverages vector-based representations of audio clips to facilitate the retrieval of sounds. This study evaluates the retrieval performances of one manually engineered audio representation and thirteen deep audio embeddings in the semantic sound similarity task. By employing a diverse range of models, our research investigates the effects of utilizing different input modalities and training objectives. In the process, we explore various design choices for integrating embeddings into sound similarity systems. Our evaluation is based on objective ranked performance metrics that incorporate sound classes and sound families, complemented by preliminary subjective assessments. We observe that the multimodal models using audio and language modalities outperform audio-only models by a significant margin, which in turn outperform audio and image models. Notably, the state-of-the-art models on the sound event classification task are not the topperforming models on the semantic sound similarity task. In addition, our findings in embedding processing methods and similarity search functions provide insights broadly applicable to information retrieval systems across different modalities.

In this work, we investigate an approach that relies on contrastive learning and music metadata as a weak source of supervision to train music representation models. Recent studies show that contrastive learning can be used with editorial metadata (e.g., artist or album name) to learn audio representations that are useful for different classification tasks. In this paper, we extend this idea to using playlist data as a source of music similarity information and investigate three approaches to generate anchor and positive track pairs. We evaluate these approaches by fine-tuning the pre-trained models for music multi-label classification tasks (genre, mood, and instrument tagging) and music similarity. We find that creating anchor and positive track pairs by relying on co-occurrences in playlists provides better music similarity and competitive classification results compared to choosing tracks from the same artist as in previous works. Additionally, our best pre-training approach based on playlists provides superior classification performance for most datasets.

An important amount of work has been devoted to the task of music classification. Despite promising results achieved by convolutional neural networks, there still exists a gap left to be filled for such models to perform well in real-world applications. In this work, we address the issue of ambiguity that can arise in many classification problems. We propose a method based on adaptive label smoothing that aims at implicitly modelling perceptual vagueness among classes to improve both training and testing performances. We assess our method using two state-of-the-art CNN architectures for audio classification on a variety of music mood and genre classification tasks. We show that the proposed strategy brings consistent improvements over the traditional approach, significantly improves generalization to external audio collections and emphasizes how crucial information carried by labels can be in an ambiguous music classification context.

One of the main limitations in the field of audio signal processing is the lack of large public datasets with audio representations and high-quality annotations due to restrictions of copyrighted commercial music. We present Melon Playlist Dataset, a public dataset of mel-spectrograms for 649,091 tracks and 148,826 associated playlists annotated by 30,652 different tags. All the data is gathered from Melon, a popular Korean streaming service. The dataset is suitable for music information retrieval tasks, in particular, auto-tagging and automatic playlist continuation. Even though the latter can be addressed by collaborative filtering approaches, audio provides opportunities for research on track suggestions and building systems resistant to the cold-start problem, for which we provide a baseline. Moreover, the playlists and the annotations included in the Melon Playlist Dataset make it suitable for metric learning and representation learning.

Modeling various aspects that make a music piece unique is a challenging task, requiring the combination of multiple sources of information. Deep learning is commonly used to obtain representations using various sources of information, such as the audio, interactions between users and songs, or associated genre metadata. Recently, contrastive learning has led to representations that generalize better compared to traditional supervised methods. In this paper, we present a novel approach that combines multiple types of information related to music using cross-modal contrastive learning, allowing us to learn an audio feature from heterogeneous data simultaneously. We align the latent representations obtained from playlists-track interactions, genre metadata, and the tracks' audio, by maximizing the agreement between these modality representations using a contrastive loss. We evaluate our approach in three tasks, namely, genre classification, playlist continuation and automatic tagging. We compare the performances with a baseline audio-based CNN trained to predict these modalities. We also study the importance of including multiple sources of information when training our embedding model. The results suggest that the proposed method outperforms the baseline in all the three downstream tasks and achieves comparable performance to the state-of-the-art.

Essentia is a reference open-source C++/Python library for audio and music analysis. In this work, we present a set of algorithms that employ TensorFlow in Essentia, allow predictions with pre-trained deep learning models, and are designed to offer flexibility of use, easy extensibility, and real-time inference. To show the potential of this new interface with TensorFlow, we provide a number of pre-trained state-of-the-art music tagging and classification CNN models. We run an extensive evaluation of the developed models. In particular, we assess the generalization capabilities in a cross-collection evaluation utilizing both external tag datasets as well as manual annotations tailored to the taxonomies of our models.

We present Essentia 2.0, an open-source C++ library for audio analysis and audio-based music information retrieval released under the Affero GPL license. It contains an extensive collection of reusable algorithms which implement audio input/output functionality, standard digital signal processing blocks, statistical characterization of data, and a large set of spectral, temporal, tonal and high-level music descriptors. The library is also wrapped in Python and includes a number of predefined executable extractors for the available music descriptors, which facilitates its use for fast prototyping and allows setting up research experiments very rapidly. Furthermore, it includes a Vamp plugin to be used with Sonic Visualiser for visualization purposes. The library is cross-platform and currently supports Linux, Mac OS X, and Windows systems. Essentia is designed with a focus on the robustness of the provided music descriptors and is optimized in terms of the computational cost of the algorithms. The provided functionality, specifically the music descriptors included in-the-box and signal processing algorithms, is easily expandable and allows for both research experiments and development of large-scale industrial applications.

Measuring music similarity is essential for multimedia retrieval. For music items, this task can be regarded as obtaining a suitable distance measurement between songs defined on a certain feature space. In this paper, we propose three of such distance measures based on the audio content: first, a low-level measure based on tempo-related description; second, a high-level semantic measure based on the inference of different musical dimensions by support vector machines. These dimensions include genre, culture, moods, instruments, rhythm, and tempo annotations. Third, a hybrid measure which combines the above-mentioned distance measures with two existing low-level measures: a Euclidean distance based on principal component analysis of timbral, temporal, and tonal descriptors, and a timbral distance based on single Gaussian Mel-frequency cepstral coefficient (MFCC) modeling. We evaluate our proposed measures against a number of baseline measures. We do this objectively based on a comprehensive set of music collections, and subjectively based on listeners' ratings. Results show that the proposed methods achieve accuracies comparable to the baseline approaches in the case of the tempo and classifier-based measures. The highest accuracies are obtained by the hybrid distance. Furthermore, the proposed classifier-based approach opens up the possibility to explore distance measures that are based on semantic notions.

Studying the ways to recommend music to a user is a central task within the music information research community. From a content-based point of view, this task can be regarded as obtaining a suitable distance measurement between songs defined on a certain feature space. We propose two such distance measures. First, a low-level measure based on tempo-related aspects, and second, a high-level semantic measure based on regression by support vector machines of different groups of musical dimensions such as genre and culture, moods and instruments, or rhythm and tempo. We evaluate these distance measures against a number of state-of-the-art measures objectively, based on 17 ground truth musical collections, and subjectively, based on 12 listeners’ ratings. Results show that, in spite of being conceptually different, the proposed methods achieve comparable or even higher performance than the considered baseline approaches. Furthermore, they open up the possibility to explore distance metrics that are based on truly semantic notions.