Mathematik und Signalverarbeitung in der Akustik

Mathematik und Signalverarbeitung in der Akustik - Laufende Projekte

French-Austrian bilateral research project funded by the French National Agency of Research (ANR) and the Austrian Science Fund (FWF, project no. I 1362-N30). The project involves two academic partners, namely the Laboratory of Mechanics and Acoustics (LMA - CNRS UPR 7051, France) and the Acoustics Research Institute. At the ARI, two research groups are involved in the project: the Mathematics and Signal Processing in Acoustics and the Psychoacoustics and Experimental Audiology groups.

Principal investigators: Thibaud Necciari (ARI), Piotr Majdak (ARI) and Olivier Derrien (LMA).

Running period: 2014-2017 (project started on March 1, 2014).

Abstract:

One of the greatest challenges in signal processing is to develop efficient signal representations. An efficient representation extracts relevant information and describes it with a minimal amount of data. In the specific context of sound processing, and especially in audio coding, where the goal is to minimize the size of binary data required for storage or transmission, it is desirable that the representation takes into account human auditory perception and allows reconstruction with a controlled amount of perceived distortion. Over the last decades, many psychoacoustical studies investigated auditory masking, an important property of auditory perception. Masking refers to the degradation of the detection threshold of a sound in presence of another sound. The results were used to develop models of either spectral or temporal masking. Attempts were made to simply combine these models to account for time-frequency (t-f) masking effects in perceptual audio codecs. We recently conducted psychoacoustical studies on t-f masking. They revealed the inaccuracy of those models which revealed the inaccuracy of such simple models. These new data on t-f masking represent a crucial basis to account for masking effects in t-f representations of sounds. Although t-f representations are standard tools in audio processing, the development of a t-f representation of audio signals that is mathematically-founded, perception-based, perfectly invertible, and possibly with a minimum amount of redundancy, remains a challenge. POTION thus addresses the following questions:

To what extent is it possible to obtain a perception-based (i.e., as close as possible to “what we see is what we hear”), perfectly invertible, and possibly minimally redundant t-f representation of sound signals? Such a representation is essential for modeling complex masking interactions in the t-f domain and is expected to improve our understanding of auditory processing of real-world sounds. Moreover, it is of fundamental interest for many audio applications involving sound analysis-synthesis.
Is it possible to improve current perceptual audio codecs by considering a joint t-f approach? To reduce the size of digital audio files, perceptual audio codecs like MP3 decompose sounds into variable-length time segments, apply a frequency transform, and use masking models to control the sub-quantization of transform coefficients within each segment. Thus, current codecs follow mainly a spectral approach, although temporal masking effects are taken into account in some implementations. By combining an efficient perception-based t-f transform with a joint t-f masking model in an audio codec, we expect to achieve significant performance improvements.

Working program:

POTION is structured in three main tasks:

Perception-based t-f representation of audio signals with perfect reconstruction: A linear and perfectly invertible t-f representation will be created by exploiting the recently developed non-stationary Gabor theory as a mathematical background. The transform will be designed so that t-f resolution mimics the t-f analysis properties by the auditory system and possibly no redundancy is introduced to maximize the coding efficiency.
Development and implementation of a t-f masking model: Based on psychoacoustical data on t-f masking collected by the partners in previous projects and on literature data, a new, complex model of t-f masking will be developed and implemented in the computationally efficient representation built in task 1. Additional psychoacoustical data required for the development of the model, involving frequency, level, and duration effects in masking for either single or multiple maskers will be collected. The resulting signal processing algorithm should represent and re-synthesize only the perceptually relevant components of the signal. It will be calibrated and validated by conducting listening tests with synthetic and real-world sounds.
Optimization of perceptual audio codecs: This task represents the main application of POTION. It will consist in combining the new efficient representation built in task 1 with the new t-f masking model built in task 2 for implementation in a perceptual audio codec.

More information on the project can be found on the POTION web page.

Publications:

Chardon, G., Necciari, Th., Balazs, P. (2014): Perceptual matching pursuit with Gabor dictionaries and time-frequency masking, in: Proceedings of the 39th International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2014). Florence, Italy, 3126-3130. (proceedings) ICASSP 2014: Perceptual matching pursuit results

Workshops:

Kick-off workshop (Vienna, 7.-8.May 2012)
Mid-term workshop (Marseille, 18.-20-November 2014)
FLAME Physics workshop (Vienna, 21.-22.Oktober 2015)

Multipliers

General description:

We live in the age of information where the analysis, classification, and transmission of information is f essential importance. Signal processing tools and algorithms form the backbone of important technologieslike MP3, digital television, mobile phones and wireless networking. Many signal processing algorithms have been adapted for applications in audio and acoustics, also taking into account theproperties of the human auditory system.

The mathematical concept of frames describes a theoretical background for signal processing. Frames are generalizations of orthonormal bases that give more freedom for the analysis and modificationof information - however, this concept is still not firmly rooted in applied research. The link between the mathematical frame theory, the signal processing algorithms, their implementations andfinally acoustical applications is a very promising, synergetic combination of research in different fields.

Therefore the main goal of this multidisciplinary project is to

-> Establish Frame Theory as Theoretical Backbone of Acoustical Modeling

in particular in psychoacoustics, phonetic and computational acoustics as well as audio engineering.

Overview

For this auspicious connection of disciplines, FLAME will produce substantial impact on both the heory and applied research.

The theory-based part of FLAME consists of the following topics:

T1 Frame Analysis and Reconstruction Beyond Classical Approaches
T2 Frame Multipliers, Extended
T3 Novel Frame Representation of Operators Motivated by Computational Acoustics

The application-oriented part of FLAME consists of:

A1 Advanced Frame Methods for Perceptual Sparsity in the Time-Frequency Plane
A2 Advanced Frame Methods for the Analysis and Classification of Speech
A3 Advanced Frame Methods for Signal Enhancement and System Estimation

Press information:

Objective:

The identification of the parameters of the vocal tract system can be used for speaker identification.

Method:

A preferred speech coding technique is the so-called Model-Based Speech Coding (MBSC), which involves modeling the vocal tract as a linear time-variant system (synthesis filter). The system's input is either white noise or a train of impulses. For coding purposes, the synthesis filter is assumed to be time-invariant during a short time interval (time slot) of typically 10-20 msec. Then, the signal is represented by the coefficients of the synthesis filter corresponding to each time slot.

A successful MBSC method is the so-called Linear Prediction Coding (LPC). Roughly speaking, the LPC technique models the synthesis filter as an all-pole linear system. This all-pole linear system has coefficients obtained by adapting a predictor of the output signal, based on its own previous samples. The use of an all-pole model provides a good representation for the majority of speech sounds. However, the representation of nasal sounds, fricative sounds, and stop consonants requires the use of a zero-pole model. Also, the LPC technique is not adequate when the voice signal is corrupted by noise.

We propose a method to estimate a zero-pole model which is able to provide the optimal synthesis filter coefficients, numerically efficient and optimal when minimizing a logarithm criterion.