This paper addresses the problem of global tempo estimation in musical audio. Given that annotating tempo is time-consuming and requires certain musical expertise, few publicly available data sources exist to train machine learning models for this task. Towards alleviating this issue, we propose a fully self-supervised approach that does not rely on any human labeled data. Our method builds on the fact that generic (music) audio embeddings already encode a variety of properties, including information about tempo, making them easily adaptable for downstream tasks. While recent work in self-supervised tempo estimation aimed to learn a tempo specific representation that was subsequently used to train a supervised classifier, we reformulate the task into the binary classification problem of predicting whether a target track has the same or a different tempo compared to a reference. While the former still requires labeled training data for the final classification model, our approach uses arbitrary unlabeled music data in combination with time-stretching for model training as well as a small set of synthetically created reference samples for predicting the final tempo. Evaluation of our approach in comparison with the state-of-the-art reveals highly competitive performance when the constraint of finding the precise tempo octave is relaxed.

Audio embeddings are crucial tools in understanding large catalogs of music. Typically embeddings are evaluated on the basis of the performance they provide in a wide range of downstream tasks, however few studies have investigated the local properties of the embedding spaces themselves which are important in nearest neighbor algorithms, commonly used in music search and recommendation. In this work we show that when learning audio representations on music datasets via contrastive learning, musical properties that are typically homogeneous within a track (e.g., key and tempo) are reflected in the locality of neighborhoods in the resulting embedding space. By applying appropriate data augmentation strategies, localisation of such properties can not only be reduced but the localisation of other attributes is increased. For example, locality of features such as pitch and tempo that are less relevant to non-expert listeners, may be mitigated while improving the locality of more salient features such as genre and mood, achieving state-of-the-art performance in nearest neighbor retrieval accuracy. Similarly, we show that the optimal selection of data augmentation strategies for contrastive learning of music audio embeddings is dependent on the downstream task, highlighting this as an important embedding design decision.

Audio embeddings enable large scale comparisons of the similarity of audio files for applications such as search and recommendation. Due to the subjectivity of audio similarity, it can be desirable to design systems that answer not only whether audio is similar, but similar in what way (e.g., wrt. tempo, mood or genre). Previous works have proposed disentangled embedding spaces where subspaces representing specific, yet possibly correlated, attributes can be weighted to emphasize those attributes in downstream tasks. However, no research has been conducted into the independence of these subspaces, nor their manipulation, in order to retrieve tracks that are similar but different in a specific way. Here, we explore the manipulation of tempo in embedding spaces as a case-study towards this goal. We propose tempo translation functions that allow for efficient manipulation of tempo within a pre-existing embedding space whilst maintaining other properties such as genre. As this translation is specific to tempo it enables retrieval of tracks that are similar but have specifically different tempi. We show that such a function can be used as an efficient data augmentation strategy for both training of downstream tempo predictors, and improved nearest neighbor retrieval of properties largely independent of tempo.

We introduce VEMOCLAP: Video EMOtion Classifier using Pretrained features, the first readily available and open-source web application that analyzes the emotional content of any user-provided video. We improve our previous work, which exploits open-source pretrained models that work on video frames and audio, and then efficiently fuse the resulting pretrained features using multi-head cross-attention. Our approach increases the state-of-the-art classification accuracy on the Ekman-6 video emotion dataset by 4.3% and offers an online application for users to run our model on their own videos or YouTube videos. We invite the readers to try our application at serkansulun.com/app.

Self-supervision methods learn representations by solving pretext tasks that do not require human-generated labels, alleviating the need for time-consuming annotations. These methods have been applied in computer vision, natural language processing, environmental sound analysis, and recently in music information retrieval, e.g. for pitch estimation. Particularly in the context of music, there are few insights about the fragility of these models regarding different distributions of data, and how they could be mitigated. In this paper, we explore these questions by dissecting a self-supervised model for pitch estimation adapted for tempo estimation via rigorous experimentation with synthetic data. Specifically, we study the relationship between the input representation and data distribution for self-supervised tempo estimation.

To model the periodicity of beats, state-of-the-art beat tracking systems use “post-processing trackers” (PPTs) that rely on several empirically determined global assumptions for tempo transition, which work well for music with a steady tempo. For expressive classical music, however, these assumptions can be too rigid. With two large datasets of Western classical piano music, namely the Aligned Scores and Performances (ASAP) dataset and a dataset of Chopin's Mazurkas (Maz-5), we report on experiments showing the failure of existing PPTs to cope with local tempo changes, thus calling for new methods. In this paper, we propose a new local periodicity-based PPT, called predominant local pulse-based dynamic programming (PLPDP) tracking, that allows for more flexible tempo transitions. Specifically, the new PPT incorporates a method called “predominant local pulses” (PLP) in combination with a dynamic programming (DP) component to jointly consider the locally detected periodicity and beat activation strength at each time instant. Accordingly, PLPDP accounts for the local periodicity, rather than relying on a global tempo assumption. Compared to existing PPTs, PLPDP particularly enhances the recall values at the cost of a lower precision, resulting in an overall improvement of F1-score for beat tracking in ASAP (from 0.473 to 0.493) and Maz-5 (from 0.595 to 0.838).

For expressive music, the tempo may change over time, posing challenges to tracking the beats by an automatic model. The model may first tap to the correct tempo, but then may fail to adapt to a tempo change, or switch between several incorrect but perceptually plausible ones (e.g., half- or double-tempo). Existing evaluation metrics for beat tracking do not reflect such behaviors, as they typically assume a fixed relationship between the reference beats and estimated beats. In this letter, we propose a new performance analysis method, called annotation coverage ratio (ACR), that accounts for a variety of possible metric-level switching behaviors of beat trackers. The idea is to derive sequences of modified reference beats of all metrical levels for every two consecutive reference beats, and compare every sequence of modified reference beats to the subsequences of estimated beats. We show via experiments on three datasets of different genres the usefulness of ACR when being utilized alongside existing metrics, and discuss the new insights that can be gained.

In this paper we present the INESC Key Detection (IKD) system which incorporates a novel method for dynamically biasing key mode estimation using the spatial displacement of beat-synchronous Tonal Interval Vectors (TIVs). We evaluate the performance of the IKD system at finding the global key on three annotated audio datasets and using three key-defining profiles. Results demonstrate the effectiveness of the mode bias in favoring either the major or minor mode, thus allowing users to fine tune this variable to improve correct key estimates on style-specific music datasets or to balance predictions across key modes on unknown input sources.

In this paper we present a system, AutoMashUpper, for making multi-song music mashups. Central to our system is a measure of “mashability” calculated between phrase sections of an input song and songs in a music collection. We define mashability in terms of harmonic and rhythmic similarity and a measure of spectral balance. The principal novelty in our approach centres on the determination of how elements of songs can be made fit together using key transposition and tempo modification, rather than based on their unaltered properties. In this way, the properties of two songs used to model their mashability can be altered with respect to transformations performed to maximize their perceptual compatibility. AutoMashUpper has a user interface to allow users to control the parameterization of the mashability estimation. It allows users to define ranges for key shifts and tempo as well as adding, changing or removing elements from the created mashups. We evaluate AutoMashUpper by its ability to reliably segment music signals into phrase sections, and also via a listening test to examine the relationship between estimated mashability and user enjoyment.

A recent trend in the field of beat tracking for musical audio signals has been to explore techniques for measuring the level of agreement and disagreement between a committee of beat tracking algorithms. By using beat tracking evaluation methods to compare all pairwise combinations of beat tracker outputs, it has been shown that selecting the beat tracker which most agrees with the remainder of the committee, on a song-by-song basis, leads to improved performance which surpasses the accuracy of any individual beat tracker used on its own. In this paper we extend this idea towards presenting a single, standalone beat tracking solution which can exploit the benefit of mutual agreement without the need to run multiple separate beat tracking algorithms. In contrast to existing work, we re-cast the problem as one of selecting between the beat outputs resulting from a single beat tracking model with multiple, diverse input features. Through extended evaluation on a large annotated database, we show that our multi-feature beat tracker can outperform the state of the art, and thereby demonstrate that there is sufficient diversity in input features for beat tracking, without the need for multiple tracking models.

This paper presents a drum transcription algorithm adjusted to the constraints of real-time audio. We introduce an instance filtering (IF) method using sub-band onset detection, which improves the performance of a system having at its core a feature-based K-nearest neighbor classifier (KNN). The architecture proposed allows for adapting different parts of the algorithm for either bass drum, snare drum or hi-hat cymbals. The open-source system is implemented in the graphic programming languages Pure Data (PD) and Max MSP, and aims to work with a large variety of drum sets. We evaluated its performance on a database of audio samples generated from a well known collection of midi drum loops randomly matched with a diverse collection of drum sets. Both of the evaluation stages, testing and validation, show an improvement in the performance when using the instance filtering algorithm.

In this paper, an approach is presented that identifies music samples which are difficult for current state-of-the-art beat trackers. In order to estimate this difficulty even for examples without ground truth, a method motivated by selective sampling is applied. This method assigns a degree of difficulty to a sample based on the mutual disagreement between the output of various beat tracking systems. On a large beat annotated dataset we show that this mutual agreement is correlated with the mean performance of the beat trackers evaluated against the ground truth, and hence can be used to identify difficult examples by predicting poor beat tracking performance. Towards the aim of advancing future beat tracking systems, we demonstrate how our method can be used to form new datasets containing a high proportion of challenging music examples.

A new probabilistic framework for beat tracking of musical audio is presented. The method estimates the time between consecutive beat events and exploits both beat and non-beat information by explicitly modeling non-beat states. In addition to the beat times, a measure of the expected accuracy of the estimated beats is provided. The quality of the observations used for beat tracking is measured and the reliability of the beats is automatically calculated. Ak-nearest neighbor regression algorithm is proposed to predict the accuracy of the beat estimates. The performance of the beat tracking system is statistically evaluated using a database of 222 musical signals of various genres. We show that modeling non-beat states leads to a significant increase in performance. In addition, a large experiment where the parameters of the model are automatically learned has been completed. Results show that simple approximations for the parameters of the model can be used. Furthermore, the performance of the system is compared with existing algorithms. Finally, a new perspective for beat tracking evaluation is presented. We show how reliability information can be successfully used to increase the mean performance of the proposed algorithm and discuss how far automatic beat tracking is from human tapping.

In this paper, we propose a method that can identify challenging music samples for beat tracking without ground truth. Our method, motivated by the machine learning method “selective sampling,” is based on the measurement of mutual agreement between beat sequences. In calculating this mutual agreement we show the critical influence of different evaluation measures. Using our approach we demonstrate how to compile a new evaluation dataset comprised of difficult excerpts for beat tracking and examine this difficulty in the context of perceptual and musical properties. Based on tag analysis we indicate the musical properties where future advances in beat tracking research would be most profitable and where beat tracking is too difficult to be attempted. Finally, we demonstrate how our mutual agreement method can be used to improve beat tracking accuracy on large music collections.

In this paper we propose an audio beat tracking system, IBT, for multiple applications. The proposed system integrates an automatic monitoring and state recovery mechanism, that applies (re-)inductions of tempo and beats, on a multi-agent-based beat tracking architecture. This system sequentially processes a continuous onset detection function while propagating parallel hypotheses of tempo and beats. Beats can be predicted in a causal or in a non-causal usage mode, which makes the system suitable for diverse applications. We evaluate the performance of the system in both modes on two application scenarios: standard (using a relatively large database of audio clips) and streaming (using long audio streams made up of concatenated clips). We show experimental evidence of the usefulness of the automatic monitoring and state recovery mechanism in the streaming scenario (i.e., improvements in beat tracking accuracy and reaction time). We also show that the system performs efficiently and at a level comparable to state-of-the-art algorithms in the standard scenario. IBT is multi-platform, open-source and freely available, and it includes plugins for different popular audio analysis, synthesis and visualization platforms.

We present a new evaluation method for measuring the performance of musical audio beat tracking systems. Central to our method is a novel visualization, the beat error histogram, which illustrates the metrical relationship between two qausi-periodic sequences of time instants: the output of beat tracking system and a set of ground truth annotations. To quantify beat tracking performance we derive an information theoretic statistic from the histogram. Results indicate that our method is able to measure performance with greater precision than existing evaluation methods and implicitly cater for metrical ambiguity in tapping sequences.

In this paper we explore the relationship between the temporal and rhythmic structure of musical audio signals. Using automatically extracted rhythmic structure we present a rhythmically-aware method to combine note onset detection techniques. Our method uses top-down knowledge of repetitions of musical events to improve detection performance by modelling the temporal distribution of onset locations. Results on a publicly available database demonstrate that using musical knowledge in this way can lead to significant improvements by reducing the number of missed and spurious detections.

Despite continued attention toward the problem of automatic beat detection in musical audio, the issue of how to evaluate beat tracking systems remains pertinent and controversial. As yet no consistent evaluation metric has been adopted by the research community. To this aim, we propose a new method for beat tracking evaluation by measuring beat accuracy in terms of the entropy of a beat error histogram. We demonstrate the ability of our approach to address several shortcomings of existing methods.

We present a simple and efficient method for beat tracking of musical audio. With the aim of replicating the human ability of tapping in time to music, we formulate our approach using a two state model. The first state performs tempo induction and tracks tempo changes, while the second maintains contextual continuity within a single tempo hypothesis. Beat times are recovered by passing the output of an onset detection function through adaptively weighted comb filterbank matrices to separately identify the beat period and alignment. We evaluate our beat tracker both in terms of the accuracy of estimated beat locations and computational complexity. In a direct comparison with existing algorithms, we demonstrate equivalent performance at significantly reduced computational cost

In this paper we apply a two state switching model to the problem of audio based beat tracking. Our analysis is based around the generation and application of adaptively weighted comb filterbank structures to extract beat timing information from the midlevel representation of an input audio signal known as the onset detection function. We evaluate our system using a previously published dataset (S. Hainsworth, Ph.D. thesis, Cambridge Univ., 2004), and in performing a direct comparison with the current state of the art, present comparable results.