- factory address
D Building,Pioneer Park,3rd Industrial Zone Fenghuang Village, Fuyong Town,Baoan DistrictShenzhen,Guangdong,China
news
TWS noise reduction new solution exploration
Bone conduction plus voice recognition
The air propagates through the microphone to generate environmental sounds such noises in the air. Bone conduction collects only the voice information of the vocal cord vibration of the person (device wearer). The main function of the algorithm is to compare the sound signals obtained from the two paths to obtain the speech sound and the environmental sound, and weaken the environmental sound, that is, finally leave a relatively clean speech sound.
AI algorithm realizes single wheat call noise reduction
TWS call noise reduction is achieved through software algorithms.
In the call noise reduction technology, deep learning is introduced into the signal processing field, and the AI noise reduction and bone conduction sensors are efficiently combined to realize the separation and extraction of vocal and background noise, covering a low SNR full scene. And the solution also requires only one microphone, which can improve engineering efficiency, save hardware cost, and achieve the first single microphone noise reduction.
The air propagates through the microphone to generate environmental sounds such noises in the air. Bone conduction collects only the voice information of the vocal cord vibration of the person (device wearer). The main function of the algorithm is to compare the sound signals obtained from the two paths to obtain the speech sound and the environmental sound, and weaken the environmental sound, that is, finally leave a relatively clean speech sound.
AI algorithm realizes single wheat call noise reduction
TWS call noise reduction is achieved through software algorithms.
In the call noise reduction technology, deep learning is introduced into the signal processing field, and the AI noise reduction and bone conduction sensors are efficiently combined to realize the separation and extraction of vocal and background noise, covering a low SNR full scene. And the solution also requires only one microphone, which can improve engineering efficiency, save hardware cost, and achieve the first single microphone noise reduction.
