A deep learning solution to the marginal stability problems of acoustic feedback systems for hearing aids

Abstract

For hearing aids, it is critical to reduce the acoustic coupling between the receiver and microphone to ensure that prescribed gains are below the maximum stable gain, thus preventing acoustic feedback. Methods for doing this include fixed and adaptive feedback cancellation, phase modulation, and gain reduction. However, the behavior of hearing aids in situations where the prescribed gain is only just below the maximum stable gain, called here “marginally stable gain,” is not well understood. This paper analyzed marginally stable systems and identified three problems, including increased gain at frequencies with the smallest gain margin, short whistles caused by the limited rate of decay of the output when the input drops, and coloration effects. A deep learning framework, called deep marginal feedback cancellation (DeepMFC), was developed to suppress short whistles, and reduce coloration effects, as well as to limit excess amplification at certain frequencies. To implement DeepMFC, many receiver signals in closed-loop systems and corresponding open-loop systems were simulated, and the receiver signals of the closed-loop and open-loop systems were paired together to obtain parallel signals for training. DeepMFC achieved much better performance than existing feedback control methods using objective and subjective measures.

Evaluation of deep marginal feedback cancellation for hearing aids using speech and music

“…DeepMFC(1) was preferred over PEM–AFC, and the difference was small but significant ( χ 2 = 10.9, p = 0.004), about 59 normal% of ratings being “equal.” For the gain margin of 0 dB, DeepMFC(1) and the combination PEM–AFC+DeepMFC(1) were equally preferred. This indicates that DeepMFC(1) alone worked well enough to make its combination with PEM–AFC unnecessary, consistent with our previous study, Zheng, Wang et al (2022). The combination PEM–AFC+DeepMFC(1) was clearly preferred over PEM–AFC.…”

“…For the gain margin of − 10 dB, the combination PEM–AFC+DeepMFC(1) was clearly preferred over DeepMFC(1) alone. Although PEM–AFC was not used in training DeepMFC(1), the results confirm that DeepMFC(1) can be used as a post-processing module to improve the performance of PEM–AFC, consistent with our previous study Zheng, Wang et al (2022).…”

“…Accordingly, the loop gain is defined as falsefalse| G false( ω false) F false( ω false) falsefalse| and the loop phase is normal∠ G false( ω false) F false( ω false). The expressions for the loop phase and loop gain, when feedback control strategies are introduced are given in Zheng, Wang et al (2022).…”

“… 3. The gain margin in decibels was defined by Zheng, Wang et al (2022) as − 20 log 10 0.2em false( K / K max false), with K prefixtruemax the maximal stable gain of the closed-loop system. When the gain margin in decibels is positive, the closed-loop system is stable in theory, but the quality of the signal is degraded when the gain margin is only slightly greater than 0 dB, as shown by Zheng, Wang et al (2022). When the gain margin in decibels is negative, the system would be unstable, that is, feedback would occur in the absence of any feedback control strategy.…”

Deep Learning-Based Road Traffic Noise Annoyance Assessment

“…With the development of deep learning, a large number of researchers have tried to use deep learning on their own tasks, and many researchers have successfully applied deep learning algorithms in many of the fields such as image segmentation [ 41 ], speech enhancement [ 42 ], hearing aids [ 43 ], traffic prediction [ 44 ], and so on. In this process, many classic deep learning model architectures have been gradually created, among which the most widely used and effective model is UNet [ 45 ]: almost all of the tasks achieved good results by using UNet architecture or modifying the UNet architecture according to the requirements of the tasks [ 41 , 44 , 45 ].…”