Fast frequency discrimination and phoneme recognition using a biomimetic membrane coupled to a neural network
This work addresses fast and accurate sound identification, potentially benefiting applications like speech processing or hearing aids, but it appears incremental as it builds on known biomimetic and neural network approaches.
The researchers tackled sound recognition by designing a biomimetic membrane that produces spatial displacement patterns in response to audible signals, which they used to train a convolutional neural network (CNN). The system outperformed existing methods like DFT, zoom FFT, and chirp z-transform in phoneme recognition, particularly in short time windows, and could unambiguously distinguish closely spaced frequency tones.
In the human ear, the basilar membrane plays a central role in sound recognition. When excited by sound, this membrane responds with a frequency-dependent displacement pattern that is detected and identified by the auditory hair cells combined with the human neural system. Inspired by this structure, we designed and fabricated an artificial membrane that produces a spatial displacement pattern in response to an audible signal, which we used to train a convolutional neural network (CNN). When trained with single frequency tones, this system can unambiguously distinguish tones closely spaced in frequency. When instead trained to recognize spoken vowels, this system outperforms existing methods for phoneme recognition, including the discrete Fourier transform (DFT), zoom FFT and chirp z-transform, especially when tested in short time windows. This sound recognition scheme therefore promises significant benefits in fast and accurate sound identification compared to existing methods.