Download A Generic System for Audio Indexing: Application to Speech/Music Segmentation and Music Genre Recognition In this paper we present a generic system for audio indexing (classification/ segmentation) and apply it to two usual problems: speech/ music segmentation and music genre recognition. We first present some requirements for the design of a generic system. The training part of it is based on a succession of four steps: feature extraction, feature selection, feature space transform and statistical modeling. We then propose several approaches for the indexing part depending of the local/ global characteristics of the indexes to be found. In particular we propose the use of segment-statistical models. The system is then applied to two usual problems. The first one is the speech/ music segmentation of a radio stream. The application is developed in a real industrial framework using real world categories and data. The performances obtained for the pure speech/ music classes problem are good. However when considering also the non-pure categories (mixed, bed) the performances of the system drop. The second problem is the music genre recognition. Since the indexes to be found are global, “segment-statistical models” are used leading to results close to the state of the art.
Download Automatic Music Detection in Television Productions This paper presents methods for the automatic detection of music within audio streams, in the fore- or background. The problem occurs in the context of a real-world application, namely, the analysis of TV productions w.r.t. the use of music. In contrast to plain speech/music discrimination, the problem of detecting music in TV productions is extremely difficult, since music is often used to accentuate scenes while concurrently speech and any kind of noise signals might be present. We present results of extensive experiments with a set of standard machine learning algorithms and standard features, investigate the difference between frame-level and clip-level features, and demonstrate the importance of the application of smoothing functions as a post-processing step. Finally, we propose a new feature, called Continuous Frequency Activation (CFA), especially designed for music detection, and show experimentally that this feature is more precise than the other approaches in identifying segments with music in audio streams.
Download The DESAM Toolbox: Spectral Analysis of Musical Audio In this paper is presented the DESAM Toolbox, a set of Matlab functions dedicated to the estimation of widely used spectral models for music signals. Although those models can be used in Music Information Retrieval (MIR) tasks, the core functions of the toolbox do not focus on any specific application. It is rather aimed at providing a range of state-of-the-art signal processing tools that decompose music files according to different signal models, giving rise to different “mid-level” representations. After motivating the need for such a toolbox, this paper offers an overview of the overall organization of the toolbox, and describes all available functionalities.
Download Fan Chirp Transformation for Music Representation In this work the Fan Chirp Transform (FChT), which provides an acute representation of harmonically related linear chirp signals, is applied to the analysis of pitch content in polyphonic music. The implementation introduced was devised to be computationally manageable and enables the generalization of the FChT for the analysis of non-linear chirps. The combination with the Constant Q Transform is explored to build a multi-resolution FChT. An existing method to compute pitch salience from the FChT is improved and adapted to handle polyphonic music. In this way a useful melodic content visualization tool is obtained. The results of a frame based melody detection evaluation indicate that the introduced technique is very promising as a front-end for music analysis.
Download A Comparison of Extended Source-Filter Models for Musical Signal Reconstruction Recently, we have witnessed an increasing use of the sourcefilter model in music analysis, which is achieved by integrating the source filter model into a non-negative matrix factorisation (NMF) framework or statistical models. The combination of the source-filter model and NMF framework reduces the number of free parameters needed and makes the model more flexible to extend. This paper compares four extended source-filter models: the source-filter-decay (SFD) model, the NMF with timefrequency activations (NMF-ARMA) model, the multi-excitation (ME) model and the source-filter model based on β-divergence (SFbeta model). The first two models represent the time-varying spectra by adding a loss filter and a time-varying filter, respectively. The latter two are extended by using multiple excitations and including a scale factor, respectively. The models are tested using sounds of 15 instruments from the RWC Music Database. Performance is evaluated based on the relative reconstruction error. The results show that the NMF-ARMA model outperforms other models, but uses the largest set of parameters.
Download Polyphonic Pitch Detection by Iterative Analysis of the Autocorrelation Function In this paper, a polyphonic pitch detection approach is presented, which is based on the iterative analysis of the autocorrelation function. The idea of a two-channel front-end with periodicity estimation by using the autocorrelation is inspired by an algorithm from Tolonen and Karjalainen. However, the analysis of the periodicity in the summary autocorrelation function is enhanced with a more advanced iterative peak picking and pruning procedure. The proposed algorithm is compared to other systems in an evaluation with common data sets and yields good results in the range of state of the art systems.
Download Latent Force Models for Sound: Learning Modal Synthesis Parameters and Excitation Functions from Audio Recordings Latent force models are a Bayesian learning technique that combine physical knowledge with dimensionality reduction — sets of coupled differential equations are modelled via shared dependence on a low-dimensional latent space. Analogously, modal sound synthesis is a technique that links physical knowledge about the vibration of objects to acoustic phenomena that can be observed in data. We apply latent force modelling to sinusoidal models of audio recordings, simultaneously inferring modal synthesis parameters (stiffness and damping) and the excitation or contact force required to reproduce the behaviour of the observed vibrational modes. Exposing this latent excitation function to the user constitutes a controllable synthesis method that runs in real time and enables sound morphing through interpolation of learnt parameters.
Download A Virtual Analog Model of the Edp Wasp VCF In this paper we present a virtual analog model of the voltagecontrolled filter used in the EDP Wasp synthesizer. This circuit is an interesting case study for virtual analog modeling due to its characteristic nonlinear and highly dynamic behavior which can be attributed to its unusual design. The Wasp filter consists of a state variable filter topology implemented using operational transconductance amplifiers (OTAs) as the cutoff-control elements and CMOS inverters in lieu of operational amplifiers, all powered by a unipolar power supply. In order to accurately model the behavior of the circuit we propose extended models for its nonlinear components, focusing particularly on the OTAs. The proposed component models are used inside a white-box circuit modeling framework to create a digital simulation of the filter which retains the interesting characteristics of the original device.
Download Dark Velvet Noise This paper proposes dark velvet noise (DVN) as an extension of the original velvet noise with a lowpass spectrum. The lowpass spectrum is achieved by allowing each pulse in the sparse sequence to have a randomized pulse width. The cutoff frequency is controlled by the density of the sequence. The modulated pulse-width can be implemented efficiently utilizing a discrete set of recursive running-sum filters, one for each unique pulse width. DVN may be used in reverberation algorithms. Typical room reverberation has a frequency-dependent decay, where the high frequencies decay faster than the low ones. A similar effect is achieved by lowering the density and increasing the pulse-width of DVN in time, thereby making the DVN suitable for artificial reverberation.
Download Real-Time Modal Synthesis of Nonlinearly Interconnected Networks Modal methods are a long-established approach to physical modeling sound synthesis. Projecting the equation of motion of a linear, time-invariant system onto a basis of eigenfunctions yields a set of independent forced, lossy oscillators, which may be simulated efficiently and accurately by means of standard time-stepping methods. Extensions of modal techniques to nonlinear problems are possible, though often requiring the solution of densely coupled nonlinear time-dependent equations. Here, an application of recent results in numerical simulation design is employed, in which the nonlinear energy is first quadratised via a convenient auxiliary variable. The resulting equations may be updated in time explicitly, thus avoiding the need for expensive iterative solvers, dense linear system solutions, or matrix inversions. The case of a network of interconnected distributed elements is detailed, along with a real-time implementation as an audio plugin.