Physical models of rigid bodies are used for sound synthesis in applications from virtual environments to music production. Traditional methods, such as modal synthesis, often rely on computationally expensive numerical solvers, while recent deep learning approaches are limited by post-processing of their results. In this work, we present a novel end-to-end framework for training a deep neural network to generate modal resonators for a given 2D shape and material using a bank of differentiable IIR filters. We demonstrate our method on a dataset of synthetic objects but train our model using an audio-domain objective, paving the way for physically-informed synthesisers to be learned directly from recordings of real-world objects.

Modelling musical performers’ individual playing styles based on audio features is important for music education, music expression analysis and music generation. In violin performance, the perception of playing styles are mainly affected by the characteristic musical timbre, which is mostly determined by performers, instruments and recording conditions. To verify if timbre features can describe a performer’s style adequately, we examine a violinist identification method based on note-level timbre feature distributions. We first apply it using solo datasets to recognise professional violinists, then use it to identify master players from commercial concerto recordings. The results show that the designed features and method work very well for both datasets. The identification accuracy with the solo dataset using MFCCs and spectral constrast features are 0.94 and 0.91 respectively. Significantly lower but promising results are reported with the concerto dataset. Results suggest that the selected timbre features can model performers’ individual playing reasonably objectively, regardless of the instrument they play.

Polyphonic vocal recordings are an inherently challenging source separation task due to the melodic structure of the vocal parts and unique timbre of its constituents. In this work, we utilise a time-domain neural network architecture re-purposed from speech separation research and modify it to separate a capella mixtures at a high sampling rate. We use four-part (soprano, alto, tenor and bass) a capella recordings of Bach Chorales and Barbershop Quartets for our experiments. Unlike current deep learning based choral separation models where the training objective is to separate constituent sources based on their class, we train our model using a permutation invariant objective. Using this we achieve state-of-the-art results for choral music separation. We introduce a novel method to estimate harmonic overlap between sung musical notes as a measure of task complexity. We also present an analysis of the impact of randomised mixing, input lengths and filterbank lengths for our task. Our results show a moderate negative correlation between the harmonic overlap of the target sources and source separation performance. We report that training our models with randomly mixed musically-incoherent mixtures drastically reduces the performance of vocal harmony separation as it decreases the average harmonic overlap presented during training.

Deep learning has become popular for many music processing tasks with Convolutional Neural Networks (CNNs) often applied. CNNs can be computationally expensive, a problem that may be alleviated through design of compact network elements or by compressing trained networks. CNNs assemble high-level structure in a hierarchical fashion, starting from small simple local patterns. On the other hand, much structure found in music spectra, such as harmonicity, is already well-defined. Both signal representations and processing methods have previously exploited such structure. We propose FifthNet, a compact neural network that is applied to the task of Automatic Chord Recognition (ACR). The compactness of FifthNet is effected through exploiting known data structure; first by arranging the network inputs according to expected data structures, then by separating processing of the semantically meaningful dimensions of the data. FifthNet is then seen to perform similar to a state-of-the-art CNN for ACR while employing only a small percentage of the parameters and computational expense used by the CNN.

Deep Learning (DL) is commonly used in music processing tasks such as Automatic Chord Recognition (ACR), for which Convolutional Neural Networks (CNNs) are popular tools. Compression of CNNs has become a research topic of interest, focused on post-pruning of learnt networks and development of less expensive network elements. CNNs assemble high level structure in data from small simple patterns. Music signals are often processed in the spectral domain where much known structure is present. We propose the FifthNet, a neural network for chroma-based ACR that incorporates known spectral structures in its design through data manipulation. We find that FifthNet is competitive with popular ACR networks while using only a small fraction of their network parameters.

An adversarial attack is an algorithm that perturbs the input of a machine learning model in an intelligent way in order to change the output of the model. An important property of adversarial attacks is transferability. According to this property, it is possible to generate adversarial perturbations on one model and apply it the input to fool the output of a different model. Our work focuses on studying the transferability of adversarial attacks in sound event classification. We are able to demonstrate differences in transferability properties from those observed in computer vision. We show that dataset normalization techniques such as z-score normalization does not affect the transferability of adversarial attacks and we show that techniques such as knowledge distillation do not increase the transferability of attacks.

The most commonly used formats for exchanging musical information today are limited in that they represent music as flat and rigid streams of events or as raw audio signals without any structural information about the content. Such files can only be listened to in a linear way and reused and manipulated in manners determined by a target application such as a Digital Audio Workstation. The publisher has no means to incorporate their intentions or understanding of the content. This article introduces an extension of the music formalism CHARM for the representation of modifiable and reusable musical content on the Web. It discusses how various kinds of multi-hierarchical graph structures together with logical constraints can be useful to model different musical situations. In particular, we focus on presenting solutions on how to interpret, navigate and schedule such structures in order for them to be played back. We evaluate the versatility of the representation in a number of practical examples created with a Web-based implementation based on Semantic Web technologies.

Recent research on piano transcription has focused primarily on note events. Very few studies have investigated pedalling techniques, which form an important aspect of expressive piano music performance. In this paper, we propose a novel method for piano sustain-pedal detection based on Convolutional Neural Networks (CNN). Inspired by different acoustic characteristics at the start (pedal onset) versus during the pedalled segment, two binary classifiers are trained separately to learn both temporal dependencies and timbral features using CNN. Their outputs are fused in order to decide whether a portion in a piano recording is played with the sustain pedal. The proposed architecture and our detection system are assessed using a dataset with frame-wise pedal on/off annotations. An average F1 score of 0.74 is obtained for the test set. The method performs better on pieces of Romantic-era composers, who intended to deliver more colours to the piano sound through pedalling techniques.

Automatic Chord Recognition (ACR) seeks to extract chords from musical signals. Recently, deep neural network (DNN) approaches have become popular for this task, being employed for feature extraction and sequence modelling. Traditionally, the most important steps in ACR were extraction of chroma features which estimate the energy in each pitch class, and pattern matching using templates or learning-based approaches. In this paper we reconsider chroma features with template matching, employing spectral reassignment chroma with synthetic spectral templates, and find experimental results comparable to those of a recent DNN-based chroma extractor.

Detecting piano pedalling techniques in polyphonic music remains a challenging task in music information retrieval. While other piano-related tasks, such as pitch estimation and onset detection, have seen improvement through applying deep learning methods, little work has been done to develop deep learning models to detect playing techniques. In this paper, we propose a transfer learning approach for the detection of sustain-pedal techniques, which are commonly used by pianists to enrich the sound. In the source task, a convolutional neural network (CNN) is trained for learning spectral and temporal contexts when the sustain pedal is pressed using a large dataset generated by a physical modelling virtual instrument. The CNN is designed and experimented through exploiting the knowledge of piano acoustics and physics. This can achieve an accuracy score of 0.98 in the validation results. In the target task, the knowledge learned from the synthesised data can be transferred to detect the sustain pedal in acoustic piano recordings. A concatenated feature vector using the activations of the trained convolutional layers is extracted from the recordings and classified into frame-wise pedal press or release. We demonstrate the effectiveness of our method in acoustic piano recordings of Chopin's music. From the cross-validation results, the proposed transfer learning method achieves an average F-measure of 0.89 and an overall performance of 0.84 obtained using the micro-averaged F-measure. These results outperform applying the pre-trained CNN model directly or the model with a fine-tuned last layer.

The expressive nature of the voice provides a powerful medium for communicating sonic ideas, motivating recent research on methods for query by vocalisation. Meanwhile, deep learning methods have demonstrated state-of-the-art results for matching vocal imitations to imitated sounds, yet little is known about how well learned features represent the perceptual similarity between vocalisations and queried sounds. In this paper, we address this question using similarity ratings between vocal imitations and imitated drum sounds. We use a linear mixed effect regression model to show how features learned by convolutional auto-encoders (CAEs) perform as predictors for perceptual similarity between sounds. Our experiments show that CAEs outperform three baseline feature sets (spectrogram-based representations, MFCCs, and temporal features) at predicting the subjective similarity ratings. We also investigate how the size and shape of the encoded layer effects the predictive power of the learned features. The results show that preservation of temporal information is more important than spectral resolution for this application.

Onset detection is a fundamental task in musical signal processing, providing information for higher level applications. Different classes of onsets can be found in musical signals, determined as being hard, or soft, by the initial energy transfer. Most onset detectors are general purpose and attempt to detect both classes of onsets, although some specifically attempt to detect soft onsets. Temporal reassignment operators related to group delay have previously been employed in onset detectors for the purposes of soft onset detection and pruning of time-frequency elements deemed to consist of vibrato. We consider the use of temporal reassignment for the detection of hard onsets and also employ the second mixed derivative of phase as a means to prune the spectral energy. Experimental validation of the proposed approach is given, showing improvements relative to state-of-the-art general purpose onset detectors for the specific tasks.

A major goal in blind source separation to identify and separate sources is to model their inherent characteristics. While most state-of-the-art approaches are supervised methods trained on large datasets, interest in non-data-driven approaches such as Kernel Additive Modelling (KAM) remains high due to their interpretability and adaptability. KAM performs the separation of a given source applying robust statistics on the time-frequency bins selected by a source-specific kernel function, commonly the K-NN function. This choice assumes that the source of interest repeats in both time and frequency. In practice, this assumption does not always hold. Therefore, we introduce a shift-invariant kernel function capable of identifying similar spectral content even under frequency shifts. This way, we can considerably increase the amount of suitable sound material available to the robust statistics. While this leads to an increase in separation performance, a basic formulation, however, is computationally expensive. Therefore, we additionally present acceleration techniques that lower the overall computational complexity.

We introduce a convolutional recurrent neural network (CRNN) for music tagging. CRNNs take advantage of convolutional neural networks (CNNs) for local feature extraction and recurrent neural networks for temporal summarisation of the extracted features. We compare CRNN with three CNN structures that have been used for music tagging while controlling the number of parameters with respect to their performance and training time per sample. Overall, we found that CRNNs show a strong performance with respect to the number of parameter and training time, indicating the effectiveness of its hybrid structure in music feature extraction and feature summarisation.

Many success stories involving deep neural networks are instances of supervised learning, where available labels power gradient-based learning methods. Creating such labels, however, can be expensive and thus there is increasing interest in weak labels which only provide coarse information, with uncertainty regarding time, location or value. Using such labels often leads to considerable challenges for the learning process. Current methods for weak-label training often employ standard supervised approaches that additionally reassign or prune labels during the learning process. The information gain, however, is often limited as only the importance of labels where the network already yields reasonable results is boosted. We propose treating weak-label training as an unsupervised problem and use the labels to guide the representation learning to induce structure. To this end, we propose two autoencoder extensions: class activity penalties and structured dropout. We demonstrate the capabilities of our approach in the context of score-informed source separation of music.

The task of chord recognition in music signals is often based upon pattern matching in chromagrams. Many variants of chroma exist and quality of chord recognition is related to the feature employed. Chroma Reduced Pitch (CRP) features are interesting in this context as they were designed to improve timbre invariance for the purpose of query retrieval. Their reapplication to chord recognition, however, has not been successful in previous studies. We consider that the default parametrisation of CRP attenuates some tonal information, as well as timbral, and consider alternatives to this default. We also provide a variant of a recently proposed compositional chroma feature, adapted for music pieces, rather than one instrument. Experiments described show improved results compared to existing features.

The estimation of rhythmic properties such as tempo, beat positions or metrical structure are central aspects of Music Information Retrieval (MIR) research. Meter inference algorithms are typically designed to track metrical structure in presence of mild deviations of the feature estimates over time in order to account for performance imprecisions, expressive timing or musical effects such as accelerando. Abrupt changes of metrical structure over time are comparatively rarely addressed. In this paper, we present an unsupervised approach to detect metrical structure changes. Formulating the problem as a metrical structure based segmentation retrieval task, we present a variant of sparse NMF and compare it to existing methods. For evaluation, we introduce a new dataset of music recordings containing metric modulations with the corresponding annotations.

In live and studio recordings unexpected sound events often lead to interferences in the signal. For non-stationary interferences, sound source separation techniques can be used to reduce the interference level in the recording. In this context, we present a novel approach combining the strengths of two algorithmic families: NMF and KAM. The recent KAM approach applies robust statistics on frames selected by a source-specific kernel to perform source separation. Based on semi-supervised NMF, we extend this approach in two ways. First, we locate the interference in the recording based on detected NMF activity. Second, we improve the kernel-based frame selection by incorporating an NMF-based estimate of the clean music signal. Further, we introduce a temporal context in the kernel, taking some musical structure into account. Our experiments show improved separation quality for our proposed method over a state-of-the-art approach for interference reduction.

This article faces the problem of how different audio features and segmentation methods work with different music genres. A new annotated corpus of Chinese traditional Jingju music is presented. We incorporate this dataset with two existing music datasets from the literature in an integrated retrieval system to evaluate existing features, structural hypotheses, and segmentation algorithms outside a Western bias. A harmonic-percussive source separation technique is introduced to the feature extraction process and brings significant improvement to the segmentation. Results show that different features capture the structural patterns of different music genres in different ways. Novelty- or homogeneity-based segmentation algorithms and timbre features can surpass the investigated alternatives for the structure analysis of Jingju due to their lack of harmonic repetition patterns. Findings indicate that the design of audio features and segmentation algorithms as well as the consideration of contextual information related to the music corpora should be accounted dependently in an effective segmentation system.

Performance of Non-negative Matrix Factorisation (NMF) can be diminished when the underlying factors consist of elements that overlap in the matrix to be factorised. The use of ℓ0sparsity may improve NMF, however such approaches are generally limited to Euclidean distance. We have previously proposed a stepwise ℓ0method for Hellinger distance, leading to improved sparse NMF. We extend sparse Hellinger NMF by proposing an alternative Iterative Hard Thresholding sparse approximation method. Experimental validation of the proposed approach is given, with a large improvement over NMF methods when learning is performed on a large dataset.

The design of systems for automatic audio feature extraction is a central aspect of the field of Music Information Retrieval. However, feature extraction systems often do not provide an indication of the reliability of the corresponding feature. Nevertheless, the provision of a reliability or confidence measure can be critical for the usage of a given feature in complex systems and real-world applications. In the present study we investigate the relationship between the entropy of a rhythmogram, which has been proposed as a descriptor of tempo salience in previous work, and the reliability of the extraction of multiple high level rhythm related features. The results show that this single descriptor is viable for simultaneously estimating the reliability of multiple rhythm features extraction. The results also provide quantitative insight that is consistent with qualitative observations extensively reported in the literature on a qualitative basis.

Similar to non-negative matrix factorization (NMF), complex matrix factorization (CMF) can be used to decompose a given music recording into individual sound sources. In contrast to NMF, CMF models both the magnitude and phase of a source, which can improve the separation of overlapped partials. However, the shift-invariance for spectral templates enabling NMF-based methods to efficiently model vibrato in music is not available with CMF. Further, the estimation of an entire phase matrix for each source results in a high number of parameters in CMF, which often leads to poor local minima. In this paper we show that score information provides a source of prior knowledge rich enough to stabilize the CMF parameter estimation, without sacrificing its expressive power. As a second contribution, we present a shift-invariant extension to CMF bringing the vibrato-modeling capabilities of NMF to CMF. As our experiments demonstrate our proposed method consistently improves the separation quality for overlapped partials compared to score-informed NMF.

This study investigates whether taking genre into account is beneficial for automatic music mood annotation in terms of core affects valence, arousal, and tension, as well as several other mood scales. Novel techniques employing genre-adaptive semantic computing and audio-based modelling are proposed. A technique called the ACTwg employs genre-adaptive semantic computing of mood-related social tags, whereas ACTwg-SLPwg combines semantic computing and audio-based modelling, both in a genre-adaptive manner. The proposed techniques are experimentally evaluated at predicting listener ratings related to a set of 600 popular music tracks spanning multiple genres. The results show that ACTwg outperforms a semantic computing technique that does not exploit genre information, and ACTwg-SLPwg outperforms conventional techniques and other genre-adaptive alternatives. In particular, improvements in the prediction rates are obtained for the valence dimension which is typically the most challenging core affect dimension for audio-based annotation. The specificity of genre categories is not crucial for the performance of ACTwg-SLPwg. The study also presents analytical insights into inferring a concise tag-based genre representation for genre-adaptive music mood analysis.

Given a musical audio recording, the goal of automatic music transcription is to determine a score-like representation of the piece underlying the recording. Despite significant interest within the research community, several studies have reported on a “glass ceiling” effect, an apparent limit on the transcription accuracy that current methods seem incapable of overcoming. In this paper, we explore how much this effect can be mitigated by focusing on a specific instrument class and making use of additional information on the recording conditions available in studio or home recording scenarios. In particular, exploiting the availability of single note recordings for the instrument in use, we develop a novel signal model employing variable-length spectro-temporal patterns as its central building blocks-tailored for pitched percussive instruments such as the piano. Temporal dependencies between spectral templates are modeled, resembling characteristics of factorial scaled hidden Markov models (FS-HMM) and other methods combining nonnegative matrix factorization with Markov processes. In contrast to FS-HMMs, our parameter estimation is developed in a global, relaxed form within the extensible alternating direction method of multipliers framework, which enables the systematic combination of basic regularizers propagating sparsity and local stationarity in note activity with more complex regularizers imposing temporal semantics. The proposed method achieves an f -measure of 93-95% for note onsets on pieces recorded on a Yamaha Disklavier (MAPS DB).

Given a musical audio recording, the goal of music transcription is to determine a score-like representation of the piece underlying the recording. Most current transcription methods employ variants of non-negative matrix factorization (NMF), which often fails to robustly model instruments producing non-stationary sounds. Using entire time-frequency patterns to represent sounds, non-negative matrix deconvolution (NMD) can capture certain types of non-stationary behavior but is only applicable if all sounds have the same length. In this paper, we present a novel method that combines the non-stationarity modeling capabilities available with NMD with the variable note lengths possible with NMF. Identifying frames in NMD patterns with states in a dynamical system, our method iteratively generates sound-object candidates separately for each pitch, which are then combined in a global optimization. We demonstrate the transcription capabilities of our method using piano pieces assuming the availability of single note recordings as training data.

Non-negative Matrix Factorisation (NMF) is a commonly used tool in many musical signal processing tasks, including Automatic Music Transcription (AMT). However unsupervised NMF is seen to be problematic in this context, and harmonically constrained variants of NMF have been proposed. While useful, the harmonic constraints may be constrictive in mixed signals. We have previously observed that recovery of overlapping signal elements using NMF is improved through introduction of a sparse coding step, and propose here the incorporation of a sparse coding step using the Hellinger distance into a NMF algorithm. Improved AMT results for unsupervised NMF are reported.

This paper presents a new set of audio features to describe music content based on tempo cues. Tempogram, a mid-level representation of tempo information, is constructed to characterize tempo variation and local pulse in the audio signal. We introduce a collection of novel tempogram-based features inspired by musicological hypotheses about the relation between music structure and its rhythmic components prominent at different metrical levels. The strength of these features is demonstrated in music structural segmentation, an important task in Music information retrieval (MIR), using several published popular music datasets. Results indicate that incorporating tempo information into audio segmentation is a promising new direction.

Techniques based on non-negative matrix factorization (NMF) have been successfully used to decompose a spectrogram of a music recording into a dictionary of templates and activations. While advanced NMF variants often yield robust signal models, there are usually some inaccuracies in the factorization since the underlying methods are not prepared for phase cancellations that occur when sounds with similar frequency are mixed. In this paper, we present a novel method that takes phase cancellations into account to refine dictionaries learned by NMF-based methods. Our approach exploits the fact that advanced NMF methods are often robust enough to provide information about how sound sources interact in a spectrogram, where they overlap, and thus where phase cancellations could occur. Using this information, the distances used in NMF are weighted entry-wise to attenuate the influence of regions with phase cancellations. Experiments on full-length, polyphonic piano recordings indicate that our method can be successfully used to refine NMF-based dictionaries.

Note onset detection and instrument recognition are two of the most investigated tasks in Music Information Retrieval (MIR). Various detection methods have been proposed in previous research for western music, with less focus on other music cultures of the world. In this paper, we focus on onset detection for percussion instruments in Beijing Opera, a major genre of Chinese traditional music. A dataset of individual audio samples of four primary percussion instruments is used to obtain the spectral bases for each instrument. With these bases, we separate the input percussion ensemble recordings into its spectral sources and their activations using a Non-negative Matrix Factorization (NMF) based algorithm. A simple onset detection conducted on each NMF activation presents satisfactory overall detection rates, and provides us valuable implications and suggestions for future development of drum transcription and percussion pattern analysis in Beijing Opera.

Mood Conductor is a system that allows the audience to interact with stage performers to create directed improvisations. The term "conductor" is used metaphorically. Rather than directing a musical performance by way of visible gestures, spectators act as conductors by communicating emotional intentions to the performers through our web-based smartphone-friendly Mood Conductor app. Performers receive the audience's directions via a visual feedback system operating in real-time. Emotions are represented by coloured blobs in a two-dimensional space (vertical dimension: arousal or excitation; horizontal dimension: valence or pleasantness). The size of the "emotion blobs" indicates the number of spectators that have selected the corresponding emotions at a given time.

The decomposition of a monaural audio recording into musically meaningful sound sources constitutes one of the central research topics in music signal processing. In this context, many recent approaches employ parametric models that describe a recording in a highly structured and musically informed way. However, a major drawback of such approaches is that the parameter learning process typically relies on computationally expensive data adaption methods. In this paper, the main idea is to distinguish parameters in which the model is linear explicitly from the remaining parameters. Exploiting the linearity we translate the data adaption problem into a sparse linear least squares problem with box constraints (SLLS-BC), a class of problems for which highly efficient numerical solvers exist. First experiments show that our approach based on modified SLLS-BC methods accelerates the data adaption by a factor of four or more compared to recently proposed methods.

We propose a subband sinusoidal (SS) synthesis algorithm for fast additive synthesis of time-varying sinusoids. The algorithm is based on a subband sinusoidal method which links a narrow band of the discrete Fourier transform (DFT) with a sinusoidal representation, facilitating direct access to sinusoidal parameters within the DFT framework. This enables the development of the proposed sinusoid synthesizer that exhibits accuracy close to direct sinusoidal modelling synthesis and efficiency comparable to conventional overlap-add with fast Fourier transforms.

In this paper we present a novel hybrid system that involves a formal method of automatic ontology generation for web-based audio signal processing applications. An ontology is seen as a knowledge management structure that represents domain knowledge in a machine interpretable format. It describes concepts and relationships within a particular domain, in our case, the domain of musical instruments. However, the different tasks of ontology engineering including manual annotation, hierarchical structuring and organization of data can be laborious and challenging. For these reasons, we investigate how the process of creating ontologies can be made less dependent on human supervision by exploring concept analysis techniques in a Semantic Web environment. In this study, various musical instruments, from wind to string families, are classified using timbre features extracted from audio. To obtain models of the analysed instrument recordings, we use K-means clustering to determine an optimised codebook of Line Spectral Frequencies (LSFs), or Mel-frequency Cepstral Coefficients (MFCCs). Two classification techniques based on Multi-Layer Perceptron (MLP) neural network and Support Vector Machines (SVM) were tested. Then, Formal Concept Analysis (FCA) is used to automatically build the hierarchical structure of musical instrument ontologies. Finally, the generated ontologies are expressed using the Ontology Web Language (OWL). System performance was evaluated under natural recording conditions using databases of isolated notes and melodic phrases. Analysis of Variance (ANOVA) were conducted with the feature and classifier attributes as independent variables and the musical instrument recognition F-measure as dependent variable. Based on these statistical analyses, a detailed comparison between musical instrument recognition models is made to investigate their effects on the automatic ontology generation system. The proposed system is general and also applicable to other research fields that are related to ontologies and the Semantic Web.

Structural segmentation of musical audio signals is one of many active areas of Music Information Retrieval (MIR) research. One aspect of this important topic which has so far received little attention though is the potential advantage to be gained by utilizing multitrack audio. This paper gives an overview of current segmentation techniques, and demonstrates that by applying a particular segmentation algorithm to multitrack data, rather than the usual case of fully mixed audio, we achieve a significant and quantifiable increase in accuracy when locating segment boundaries. Additionally, we provide details of a structurally annotated multitrack test set available for use by other researchers.

Music collections are often visualized in a two-dimensional space to show relationships between songs. Some user inter faces interacting with these two-dimensional maps of songs use spatial auditory display to allow easier access to the au dio content. A common auditory display is to have multiple songs playing simultaneously from differing spatial locations around the user. However, when using this style of interface the sonification of the collection needs to be limited to a lo cal subset of the collection, usually three to six songs. This paper presents the amblr, a spatial audio music browser that allows a user to auralize the collection. It combines effective design from previous work with new approaches to create a novel interface. This allows for a more intuitive navigation of a virtual space populated by a large collection songs without relying on textual metadata.

This paper presents an extensive analysis of a sample of a social network of musicians. The network sample is first analyzed using standard complex network techniques to verify that it has similar properties to other web-derived complex networks. Content-based pairwise dissimilarity values between the musical data associated with the network sample are computed, and the relationship between those content-based distances and distances from network theory explored. Following this exploration, hybrid graphs and distance measures are constructed, and used to examine the community structure of the artist network. Finally, results of these investigations are shown to be mostly orthogonal between these distance spaces. These results are considered with a focus recommendation and discovery applications employing these hybrid measures as their basis.

This paper introduces a new database of room impulse responses. This database differs greatly from previously released databases as it contains over 700 impulse responses. The impulse responses are measured in three different rooms each with a static source position and at least 130 different receiver positions. Each measurement position is recorded with both an omnidirectional microphone and a B-format microphone.

Sonic Visualiser is a friendly and flexible end-user desktop application for analysis, visualisation, and annotation of music audio files. Its stated goal is to be "the first program you reach for when want to study a musical recording rather than simply listen to it". To this end, it has a user interface that resembles familiar audio editing applications, a set of useful standard visualisation facilities, and support for a plugin format for additional automated analysis methods.

In this paper we describe a novel approach to applying text-based information retrieval techniques to music collections. We represent tracks with a joint vocabulary consisting of both conventional words, drawn from social tags, and audiomuswords, representing characteristics of automatically-identified regions of interest within the signal. We build vector space and latent aspect models indexing words and muswords for a collection of tracks, and show experimentally that retrieval with these models is extremely well-behaved. We find in particular that retrieval performance remains good for tracks by artists unseen by our models in training, and even if tags for their tracks are extremely sparse.

We describe a method of segmenting musical audio into structural sections based on a hierarchical labeling of spectral features. Frames of audio are first labeled as belonging to one of a number of discrete states using a hidden Markov model trained on the features. Histograms of neighboring frames are then clustered into segment-types representing distinct distributions of states, using a clustering algorithm in which temporal continuity is expressed as a set of constraints modeled by a hidden Markov random field. We give experimental results which show that in many cases the resulting segmentations correspond well to conventional notions of musical form. We show further how the constrained clustering approach can easily be extended to include prior musical knowledge, input from other machine approaches, or semi-supervision.

Four music segmentation algorithms are presented, one based on purely timbral features, one on purely harmonic features, and two on different combinations of features. They are compared against each other and against human annotations of two albums by The Beatles. Example segmentations are given together with a quantitative measure of boundary accuracy. No algorithm is found to be clearly superior, although examples suggest that the combined algorithms can offer improved boundary detection.

We propose a novel system able to modify the rhythm of a given drum loop, known as the original, to match the rhythmic pattern of a second loop, known as the model. Our approach is fully automated, thus eliminating the need for MIDI sequencing. The presented methodology combines standard and state-of-the-art techniques for the segmentation and classification of drum sounds, the matching of drum sequences, and the transformation of the original loop. We discuss the advantages and disadvantages of the proposed approach and provide links to examples for the qualitative evaluation of the system's output

This paper addresses the issue of drum sound classification in the context of automatic rhythm modification of drum loops. The proposed method segments the signal using an onset detection algorithm, characterises segmented sounds using a spectral feature set, and classifies them using k-means clustering. We propose a simple taxonomy for the grouping of different instrumental sounds under a few utilitarian labels. Results demonstrate the adequacy of our proposed taxonomy while showing that our classification approach outperforms commonly-used supervised learning techniques

This paper deals with musical instrument identification. The proposed method consists of building the instrument models using a set of linear predictive coefficients, the Line Spectrum Frequencies. The models consist of characteristic short-term spectral envelopes calculated for each instrument in the database. The identification process involves the calculation of a similarity measure between two codebooks, one taken from the models database, one corresponding to the sample to identify. Next, the use of Support Vector Machines as classifier is investigated. The two systems are then applied to the identification of monophonic phrases extracted from commercial recordings. It is shown that good performance can be achieved for the classification of one unknown excerpt amongst 6 instruments.

A method for segmenting musical audio with a hierarchical timbre model is introduced. New evidence is presented to show that music segmentation can be recast as clustering of timbre features, and a new clustering algorithm is described. A prototype thumbnail-generating application is described and evaluated. Experimental results are given, including comparison of machine and human segmentations

This paper describes a method for automatically segmenting and labelling sections in recordings of musical audio. We incorporate the user's expectations for segment duration as an explicit prior probability distribution in a Bayesian framework, and demonstrate experimentally that this method can produce accurate labelled segmentations for popular music

The aim of this paper is to propose solutions to some problems that arise in automatic polyphonic transcription of recorded piano music. First, we propose a method that groups spectral information in the frequency-domain and uses a rule-based framework to deal with the known problems of polyphony and harmonicity. Then, we present a novel method for multipitch-estimation that uses both frequency and time-domain information. It assumes signal segments to be the linearly weighted sum of waveforms in a database of individual piano notes. We propose a solution to the problem of generating those waveforms, by using the frequency-domain approach. We show that accurate time-domain transcription can be achieved given an adequate estimation of the database. This suggests an alternative to common frequency-domain approaches that does not require any prior training on a separate database of isolated notes

Note onset detection and localization is useful in a number of analysis and indexing techniques for musical signals. The usual way to detect onsets is to look for "transient" regions in the signal, a notion that leads to many definitions: a sudden burst of energy, a change in the short-time spectrum of the signal or in the statistical properties, etc. The goal of this paper is to review, categorize, and compare some of the most commonly used techniques for onset detection, and to present possible enhancements. We discuss methods based on the use of explicitly predefined signal features: the signal's amplitude envelope, spectral magnitudes and phases, time-frequency representations; and methods based on probabilistic signal models: model-based change point detection, surprise signals, etc. Using a choice of test cases, we provide some guidelines for choosing the appropriate method for a given application.

Electronic Music Distribution is in need of robust and automatically extracted music descriptors. An important attribute of a piece of polyphonic music is what is commonly referred to as "the way it sounds". While there has been a large quantity of research done to model the timbre of individual instruments, little work has been done to analyze "real world" timbre mixtures such as the ones found in popular music. In this paper, we present our research about such "polyphonic timbres". We describe an effective way to model the textures found in a given music signal, and show that such timbre models provide new solutions to many issues traditionally encountered in music signal processing and music information retrieval. Notably, we describe their applications for music similarity, segmentation and pattern induction.

We present a study on the combined use of energy and phase information for the detection of onsets in musical signals. The resulting method improves upon both energy-based and phase-based approaches. The detection function, generated from the analysis of the signal in the complex frequency domain is sharp at the position of onsets and smooth everywhere else. Results on a database of recordings show high detection rates for low rates of errors. The approach is more robust than its predecessors both theoretically and practically.

Although widely used in many coding applications, the Modified Discrete Cosine Transform (MDCT) has the drawback of being sensitive to time shifts. With the popular choice of a sine window, we show that it is possible to compute an explicit formulation of this time dependency. Starting from the exact MDCT of a pure sine and a simple interpretation in terms of combined modulations, we propose a regularization method that computes a "pseudo-spectrum" for the set of MDCT coefficients. This pseudo-spectrum is shown to provide, at a low computational cost, a good approximation of the local spectrum of the signal, with an improved behavior with respect to frequency and phase than the classical MDCT spectrum, ie. the absolute value of the coefficients. Amongst other applications, this procedure can be used to reduce some of the artifacts that appear in MDCT-based audio coders at low bit-rates.

Note onsets mark the beginning of attack transients, short areas of a note containing rapid changes of the signal spectral content. Detecting onsets is not trivial, especially when analysing complex mixtures. Applications for note onset detection systems include time stretching, audio coding and synthesis. An alternative to standard energy-based onset detection is proposed by using phase information. It is suggested that by observing the frame-by-frame distribution of differential angles, the precise moment when onsets occur can be detected with accuracy. Statistical measures are used to build the detection function. The system is tested and tuned on a database of complex recordings.

In this article, we give an overview of a range of approaches to the analysis and separation of musical audio. In particular, we consider the problems of automatic music transcription and audio source separation, which are of particular interest to our group. Monophonic music transcription, where a single note is present at one time, can be tackled using an autocorrelation-based method. For polyphonic music transcription, with several notes at any time, other approaches can be used, such as a blackboard model or a multiple-cause/sparse coding method. The latter is based on ideas and methods related to independent component analysis (ICA), a method for sound source separation.

This paper presents a method for evaluating the performance of multi-channel sound reproduction systems serving large audience areas. The technique, based on the synthesis of binaural cues, emulates the pressures at the listener's ears corresponding to those produced by a real sound field thus, in theory, allowing the assessment of the imaging produced by holographic sound systems to be carried out simply by using a pair of headphones. Preliminary results obtained by applying this method to a simple surround sound system are presented and discussed.

This paper presents a study on musical signal classification, using wavelet transform analysis in conjunction with statistical pattern recognition techniques. A comparative evaluation between different wavelet analysis architectures in terms of their classification ability, as well as between different classifiers is carried out. We seek to establish which statistical measures clearly distinguish between the three different musical styles of rock, piano, and jazz. Our preliminary results suggest that the features collected by the adaptive splitting wavelet transform technique performed better compared to the other wavelet based techniques, achieving an overall classification accuracy of 91.67%, using either the minimum distance classifier or the least squares minimum distance classifier. Such a system can play a useful part in multimedia applications which require content based search, classification, and retrieval of audio signals, as defined in MPEG-7.

The performance obtainable with four-tap wavelet filters for low bit rate audio coding is presented. For the investigation and comparison of the performance of these wavelet filters, a codec model has been designed and implemented based on wavelet packet algorithm and the model of auditory perception.

The performance of some different wavelet families, including for comparison a well known family of QMFs, is investigated for low bit rate coding of audio signals. For the assessment of the coding gain of these wavelets, both octave and uniform subband coding schemes have been evaluated, using both constant and dynamic bit allocation, with and without entropy noiseless Huffman coding. The influence of complexity of these wavelets, in terms of number of filter coefficients, against the quality of the decompressed audio signals in terms of segmental-SNR (dB), is presented, at different bit rates. In addition, this evaluation suggests that perceptually transparent quality of monophonic signals can be achieved at 24 kbits/sec (Fs=8 kHz, 3 bits/sample) for speech applications and at 64 kbits/sec (Fs=48 kHz, 1.33 bits/sample) for music related applications, as in digital audio transmission and storage.

Recent work has shown that high-order single-bit sigma-delta modulators suffer from low-level artifacts such as idle tones and noise modulation. Techniques that have been proposed to reduce or eliminate these errors include the application of dither inside the one-bit quantiser loop, and selecting a loop filter which makes the modulator chaotic. This paper compares the efficacy of these two approaches by simulating high-resolution sigma-delta modulators suitable for audio-conversion applications. Dynamic-range penalties for successful linearisation are determined for two types of dither signal and two classes of chaos.

There has recently been some interest in the emerging technique of Sigma Delta Signal Processing (SDSP) where direct processing of the bitstream is proposed as an alternative to the decimation-PCM processing-interpolation cycle required for DSP of /spl Sigma//spl Delta/ converted signals. SDSP requires digital (re)modulation with a variety of wordlengths in the oversampled datastream; the prediction and control of the remodulator complexity is crucial in determining the relative value of SDSP in any given context. This paper addresses this issue in deriving a new method for reducing remodulator complexity and in obtaining its dependence on the datastream wordlength.

This paper is about exploiting chaos theory to provide a means for modelling sound on a computer. A new technique is described for the analysis and resynthesis of a sound using a chaotic system. Preliminary results are promising: a synthetic version of 'air noise' is generated which sounds very much like the original.>

This paper describes the problems of numerical implementations of the IPFE algorithm for partitioning high order filters into parallel interconnected subfilters. To address these problems, a modification to the original IPFE algorithm is proposed. Although the new algorithm was developed to give improved reliability over the original algorithm, initial results indicate that due to the ill-conditioning of a key matrix, the new algorithm is much worse.>

A Hough transform based neural network shape recognition system has been developed. The authors propose a new technique for 2-D shape recognition based on the peak and envelope vectors extracted from the Hough space. The vectors are fed into a neural network which performs both the learning and recognition tasks.>

DPSs (digital signal processors), conventional microprocessors, and transputers are compared on the basis of their ability to perform image processing tasks, with a view to assessing their potential for implementation in real time systems. The study covers a selection of the more important instructions in use in image processing algorithms as well as a selection of the algorithms themselves as a simple benchmark, and includes a survey of multiprocessor systems implemented with some of these processors. It is concluded that DSPs are better suited to embedded systems, but the transputer may offer the advantage in a research environment of general usefulness for supercomputing power. Transputer systems are also more flexible in the sense of re-programmability, and a single system could be used for all processing levels, numeric and symbolic, in a computer vision system.>