Emmanouil Benetos

Detection and Classification of Acoustic Scenes and Events

For intelligent systems to make best use of the audio modality, it is important that they can recognize not just speech and music, which have been researched as specific tasks, but also general sounds in everyday environments. To stimulate research in this field we conducted a public research challenge: the IEEE Audio and Acoustic Signal Processing Technical Committee challenge on Detection and Classification of Acoustic Scenes and Events (DCASE). In this paper, we report on the state of the art in automatically classifying audio scenes, and automatically detecting and classifying audio events. We survey prior work as well as the state of the art represented by the submissions to the challenge from various research groups. We also provide detail on the organization of the challenge, so that our experience as challenge hosts may be useful to those organizing challenges in similar domains. We created new audio datasets and baseline systems for the challenge; these, as well as some submitted systems, are publicly available under open licenses, to serve as benchmarks for further research in general-purpose machine listening.

Automatic music transcription: challenges and future directions

Automatic music transcription is considered by many to be a key enabling technology in music signal processing. However, the performance of transcription systems is still significantly below that of a human expert, and accuracies reported in recent years seem to have reached a limit, although the field is still very active. In this paper we analyse limitations of current methods and identify promising directions for future research. Current transcription methods use general purpose models which are unable to capture the rich diversity found in music signals. One way to overcome the limited performance of transcription systems is to tailor algorithms to specific use-cases. Semi-automatic approaches are another way of achieving a more reliable transcription. Also, the wealth of musical scores and corresponding audio data now available are a rich potential source of training data, via forced alignment of audio to scores, but large scale utilisation of such data has yet to be attempted. Other promising approaches include the integration of information from multiple algorithms and different musical aspects.

Detection and classification of acoustic scenes and events: An IEEE AASP challenge

This paper describes a newly-launched public evaluation challenge on acoustic scene classification and detection of sound events within a scene. Systems dealing with such tasks are far from exhibiting human-like performance and robustness. Undermining factors are numerous: the extreme variability of sources of interest possibly interfering, the presence of complex background noise as well as room effects like reverberation. The proposed challenge is an attempt to help the research community move forward in defining and studying the aforementioned tasks. Apart from the challenge description, this paper provides an overview of systems submitted to the challenge as well as a detailed evaluation of the results achieved by those systems.

An end-to-end neural network for polyphonic piano music transcription

We present a supervised neural network model for polyphonic piano music transcription. The architecture of the proposed model is analogous to speech recognition systems and comprises an acoustic model and a music language model. The acoustic model is a neural network used for estimating the probabilities of pitches in a frame of audio. The language model is a recurrent neural network that models the correlations between pitch combinations over time. The proposed model is general and can be used to transcribe polyphonic music without imposing any constraints on the polyphony. The acoustic and language model predictions are combined using a probabilistic graphical model. Inference over the output variables is performed using the beam search algorithm. We perform two sets of experiments. We investigate various neural network architectures for the acoustic models and also investigate the effect of combining acoustic and music language model predictions using the proposed architecture. We compare performance of the neural network-based acoustic models with two popular unsupervised acoustic models. Results show that convolutional neural network acoustic models yield the best performance across all evaluation metrics. We also observe improved performance with the application of the music language models. Finally, we present an efficient variant of beam search that improves performance and reduces run-times by an order of magnitude, making the model suitable for real-time applications.

A shift-invariant latent variable model for automatic music transcription

In this work, a probabilistic model for multiple-instrument automatic music transcription is proposed. The model extends the shift-invariant probabilistic latent component analysis method, which is used for spectrogram factorization. Proposed extensions support the use of multiple spectral templates per pitch and per instrument source, as well as a time-varying pitch contribution for each source. Thus, this method can effectively be used for multiple-instrument automatic transcription. In addition, the shift-invariant aspect of the method can be exploited for detecting tuning changes and frequency modulations, as well as for visualizing pitch content. For note tracking and smoothing, pitch-wise hidden Markov models are used. For training, pitch templates from eight orchestral instruments were extracted, covering their complete note range. The transcription system was tested on multiple-instrument polyphonic recordings from the RWC database, a Disklavier data set, and the MIREX 2007 multi-F0 data set. Results demonstrate that the proposed method outperforms leading approaches from the transcription literature, using several error metrics.

Musical Instrument Classification using Non-Negative Matrix Factorization Algorithms and Subset Feature Selection

In this paper, a class, of algorithms for automatic classification of individual musical instrument sounds is presented. Several perceptual features used in sound classification applications as well as MPEG-7 descriptors were measured for 300 sound recordings consisting of 6 different musical instrument classes. Subsets of the feature set are selected using branch-and-bound search, obtaining the most suitable features for classification, A class of classifiers is developed based on the non-negative matrix factorization (NMF). The standard NMF method is examined as well as its modifications: the local, the sparse, and the discriminant NMF. The experimental results compare feature subsets of varying sizes alongside the various NMF algorithms. It has been found that a subset containing the mean and die variance of the first mel-frequency cepstral coefficient and the audiospectrumflatness descriptor along with the means of the audiospectrumenvelope and the audiospectrumspread descriptors when is fed to a standard NMF classifier yields an accuracy exceeding 95%

Multiple-instrument polyphonic music transcription using a temporally constrained shift-invariant model

A method for automatic transcription of polyphonic music is proposed in this work that models the temporal evolution of musical tones. The model extends the shift-invariant probabilistic latent component analysis method by supporting the use of spectral templates that correspond to sound states such as attack, sustain, and decay. The order of these templates is controlled using hidden Markov model-based temporal constraints. In addition, the model can exploit multiple templates per pitch and instrument source. The shift-invariant aspect of the model makes it suitable for music signals that exhibit frequency modulations or tuning changes. Pitch-wise hidden Markov models are also utilized in a postprocessing step for note tracking. For training, sound state templates were extracted for various orchestral instruments using isolated note samples. The proposed transcription system was tested on multiple-instrument recordings from various datasets. Experimental results show that the proposed model is superior to a non-temporally constrained model and also outperforms various state-of-the-art transcription systems for the same experiment.

Joint Multi-Pitch Detection Using Harmonic Envelope Estimation for Polyphonic Music Transcription

In this paper, a method for automatic transcription of music signals based on joint multiple-F0 estimation is proposed. As a time-frequency representation, the constant-Q resonator time-frequency image is employed, while a novel noise suppression technique based on pink noise assumption is applied in a preprocessing step. In the multiple-F0 estimation stage, the optimal tuning and inharmonicity parameters are computed and a salience function is proposed in order to select pitch candidates. For each pitch candidate combination, an overlapping partial treatment procedure is used, which is based on a novel spectral envelope estimation procedure for the log-frequency domain, in order to compute the harmonic envelope of candidate pitches. In order to select the optimal pitch combination for each time frame, a score function is proposed which combines spectral and temporal characteristics of the candidate pitches and also aims to suppress harmonic errors. For postprocessing, hidden Markov models (HMMs) and conditional random fields (CRFs) trained on MIDI data are employed, in order to boost transcription accuracy. The system was trained on isolated piano sounds from the MAPS database and was tested on classic and jazz recordings from the RWC database, as well as on recordings from a Disklavier piano. A comparison with several state-of-the-art systems is provided using a variety of error metrics, where encouraging results are indicated.

Non-Negative Tensor Factorization Applied to Music Genre Classification

Music genre classification techniques are typically applied to the data matrix whose columns are the feature vectors extracted from music recordings. In this paper, a feature vector is extracted using a texture window of one sec, which enables the representation of any 30 sec long music recording as a time sequence of feature vectors, thus yielding a feature matrix. Consequently, by stacking the feature matrices associated to any dataset recordings, a tensor is created, a fact which necessitates studying music genre classification using tensors. First, a novel algorithm for non-negative tensor factorization (NTF) is derived that extends the non-negative matrix factorization. Several variants of the NTF algorithm emerge by employing different cost functions from the class of Bregman divergences. Second, a novel supervised NTF classifier is proposed, which trains a basis for each class separately and employs basis orthogonalization. A variety of spectral, temporal, perceptual, energy, and pitch descriptors is extracted from 1000 recordings of the GTZAN dataset, which are distributed across 10 genre classes. The NTF classifier performance is compared against that of the multilayer perceptron and the support vector machines by applying a stratified 10-fold cross validation. A genre classification accuracy of 78.9% is reported for the NTF classifier demonstrating the superiority of the aforementioned multilinear classifier over several data matrix-based state-of-the-art classifiers.

Polyphonic music transcription using note onset and offset detection

In this paper, an approach for polyphonic music transcription based on joint multiple-F0 estimation and note onset/offset detection is proposed. For preprocessing, the resonator time-frequency image of the input music signal is extracted and noise suppression is performed. A pitch salience function is extracted for each frame along with tuning and inharmonicity parameters. For onset detection, late fusion is employed by combining a novel spectral flux-based feature which incorporates pitch tuning information and a novel salience function-based descriptor. For each segment defined by two onsets, an overlapping partial treatment procedure is used and a pitch set score function is proposed. A note offset detection procedure is also proposed using HMMs trained on MIDI data. The system was trained on piano chords and tested on classic and jazz recordings from the RWC database. Improved transcription results are reported compared to state-of-the-art approaches.