The algorithm of the hearing enhance earphone basically follows the idea and framework of the hearing aid, which can be considered as a dimensionality reduction of the hearing aid algorithm.
The biggest difference between TWS earphones, or general earphones, and hearing aids/hearing aid earphones is in the usage scenarios of conversations.
Headphone dialogue: the speaker is local, and the listener is far away, such as one in Beijing and the other in Shenzhen, thousands of miles away.
Hearing aid/Auditory conversation: the speaker is local, and the listener is also local, such as in a shared meal or meeting, only a few meters away.
(1) Segment gain algorithm
The official name of the segment gain algorithm is the multi-channel loudness compensation algorithm. Since the sensitivity of hearing loss patients to sound varies with frequency, digital hearing aids should set different gains for sound signals in different frequency regions, which are called channels or frequency bands. The final output signal is the synthesis of different channels after amplification. The multi-channel loudness compensation algorithm can accurately match the hearing loss of the hearing-impaired person.
In the multi-channel loudness compensation algorithm, the performance of frequency band division and synthesis is an important factor affecting the performance of the algorithm.
(2) Dialogue noise reduction algorithm
In face-to-face conversation scenarios, how to reduce background noise and improve speech intelligibility without distortion is one of the most challenging tasks in hearing aid design. Noise reduction algorithms can improve the signal-to-noise ratio and are an important method to improve patient intelligibility. However, noise in reality is often ever-changing, and useful speech and noise may overlap in space or time, making it challenging to use noise reduction algorithms to improve speech comprehension of hearing-impaired patients in noisy environments.
(3) Howling suppression algorithm
Howling suppression algorithm, or echo suppression algorithm. The gain of a digital hearing aid determines its sound amplification properties. Since the hearing aid microphone is very close to the receiver, the sound output by the receiver often leaks into the microphone through a certain path. Echoes are formed when the internal gain is large, and howling is caused in severe cases. One of the most serious problems in hearing aids.
(4) Frequency shift algorithm
The frequency shift algorithm is also called the frequency reduction algorithm. The main method of the frequency reduction algorithm is to transfer or compress the high-frequency information to the low-frequency frequency that the patient can hear, and then through language training, so that the patient can re-establish language perception habits, and then achieve the purpose of understanding language.
The complexity of the algorithm is one of the difficulties of hearing aid algorithms. For example, for the howling suppression algorithm, the digital hearing aid can adaptively estimate the echo path through the echo suppression algorithm, and then subtract the echo signal from the original signal, so as to eliminate the echo and improve the maximum gain of the hearing aid. However, the frequency spectrum of the speech signal and the noise signal is highly coincident, and the desired signal and the echo signal are correlated signals, which makes the estimation of the echo path quite complicated. How to design an algorithm to obtain an unbiased estimate of the echo path is the current digital Difficulties in the study of speech processing algorithms for hearing aids.
The engineering difficulty is mainly due to the computational power limitations of the digital signal processor (DSP) in the headset. The computing power of DSP is only one thousandth of that of server. This has led to a significant simplification of requirements for DSP algorithms compared to general cloud algorithms. It is a very big engineering challenge to maintain the validity of the speech processing results while simplifying the model.
Shenzhen WinTo Technology Co., Limited have been dedicated to develop and produce hearing enhance headphones since 2019.
Welcome to contact with us.