Koen Eneman

DFT modulated filter bank design for oversampled subband systems

Filter banks are widely used in digital signal processing, often integrated in a multirate scheme, to reduce the implementation cost and to improve algorithmic performance. DFT modulated filter banks are commonly used to design oversampled subband schemes. Oversampled subband schemes are attractive as they trade off between complexity gain and aliasing distortion. Ideally, the overall effect of a subband system is that of a pure delay, i.e. the subband system is perfectly reconstructing. In this paper we present design techniques for perfect and nearly perfect reconstruction oversampled DFT modulated filter banks. One of the disadvantages of a perfect reconstruction filter bank design is that the stopband attenuation of the filters is typically small, which could have a negative impact on intermediate subband operations such as subband adaptive filtering. Nearly perfect reconstruction filter banks can be designed to overcome this problem, leading to a better performance of the intermediate algorithms.

Multiple beam broadband beamforming: filter design and real-time implementation

In this paper the design and implementation of a truly broadband beamformer is described. Using only a limited number of microphones a beamformer has been developed with a reasonably homogeneous beampattern over the entire speech bandwidth. The solution is obtained as the result of a filter-and-sum operation. Beams are steered towards different listening-directions and for each direction a set of spatio-temporal filters had to be designed. All beams operate in parallel and a beam-selection procedure picks out the beam with the most prominent speech signal. The complete system has been implemented in real-time using 1 MOTOROLA-56001 and 4 TMS-320C40 processors.

Multimicrophone Speech Dereverberation: Experimental Validation

Dereverberation is required in various speech processing applications such as handsfree telephony and voice-controlled systems, especially when signals are applied that are recorded in a moderately or highly reverberant environment. In this paper, we compare a number of classical and more recently developed multimicrophone dereverberation algorithms, and validate the different algorithmic settings by means of two performance indices and a speech recognition system. It is found that some of the classical solutions obtain a moderate signal enhancement. More advanced subspace-based dereverberation techniques, on the other hand, fail to enhance the signals despite their high-computational load.

Hybrid subband/frequency-domain adaptive systems

Abstract For many years now, subband and frequency-domain adaptive filtering techniques have been proposed for the identification of high-order FIR systems. Classical LMS-based algorithms are less attractive as their computational load is higher and their convergence behaviour for coloured inputs is worse. Subband processing has many desirable properties. However, when used to implement adaptive filters, various effects (such as residual errors and slow convergence due to aliasing) occur and reduce performance. On the other hand, it is known that frequency-domain adaptive filters do not suffer from these problems despite being (nearly) equivalent to subband adaptive filters, be it with “poor” filter banks (modulated sinc frequency responses). In this paper the operation of the frequency-domain techniques are unravelled and explained in the “subband jargon” and among other things it will be shown how aliasing is compensated for in the frequency-domain approach. This paper aims at generalising frequency-domain aliasing-compensation techniques to subband adaptive systems. Further, some design criteria for subband adaptive systems are formulated. Three realisation conditions and an alternative adaptation scheme will be proposed. Standard subband adaptive filters cannot fulfil all conditions, whereas for frequency-domain-based algorithms the design criteria are met.

Filter bank constraints for subband and frequency-domain adaptive filters

For many years now, subband and frequency-domain adaptive filtering techniques have been proposed for the cancellation of long acoustic echoes. Classical LMS based algorithms are less attractive as their computation load is higher and the convergence behaviour for coloured far-end inputs is worse. We specify 3 realization conditions for DFT modulated subband schemes. Standard subband adaptive filters cannot fulfil all conditions. We show that frequency-domain based algorithms can be considered as a special case of subband adaptive filtering and that the realization conditions can be fulfilled in this case.

The use of virtual acoustics in the evaluation and development of binaural hearing aid algorithms

The development of noise reduction algorithms for hearing aids (HA) is not longer only related to the improvement of signal to noise ratio, but also to the quality of hearing, e.g. binaural aspects of hearing. This is very important for the recognition of the localization of sound sources but also for an improved speech intelligibility in noisy situations due to spatial release from masking eects. New design and signal processing algorithms for binaural HA’s need to be tested and validated in dierent acoustical scenarios. As it is too laborious and time consuming to perform sucient numbers of perceptual evaluations in dierent rooms with dierent acoustical parameters, advanced acoustic modeling of dierent virtual acoustical environments might be needed. Virtual acoustics in our research relates to the convolution of the measured or simulated binaural signals (head related transfer functions - HRTF’s) with the impulse response generated from a computer model of a room (using ODEON R ∞ software) to simulate binaural sounds. This study investigates the usage of virtual acoustics in the framework of developing algorithms for binaural hearing aids. It evaluates and quantifies the fidelity of binaural signals generated by commercially available virtual acoustics software with respect to the localization of sound and speech intelligibility in dierent acoustical scenarios.

Signal processing in Hearing Aids: Results of the HEARCOM project

Digital hearing aids of today allow the application of advanced signal processing strategies. In recent years a number of promising signal processing approaches have been designed and developed. However, most of these different evolutions have been evaluated only in a limited way. Within the framework of the HEARCOM EU‐research project a number of signal enhancement techniques have been further developed and evaluated based on a representative set of real‐life recordings and physical performance measures. Different auditory profiles, representing common categories of hearing aid users, have been taken into account. A selection of 5 of these signal enhancement techniques (single‐channel noise suppression, blind source separation, dereverberation, multi‐microphone adaptive processing, feedback reduction) has been implemented on a single common hard‐ and software test platform, the Master Hearing Aid (MHA). These signal processing strategies have been evaluated perceptually based on speech reception threshol...