Sound returning function component, sound insulation feedback earphone and application thereof, and sound insulation feedback method

Abstract

The invention discloses a sound returning function component, a sound insulation feedback earphone and an application thereof, and a sound insulation feedback method. The sound returning function component comprises a uni-diectivity microphone receiving head (1), a sound amplifier (2), an AD conversion module (3), a processor (4), a DA conversion module (5), and an audio operational amplifier (6)that are sequentially connected. The invention can realize that only the voice of the user is returned in a noisy environment; in addition, through loading the sound returning function component disclosed by the invention into an ordinary earphone, a plurality of students can simultaneously practice speaking or reading in the same classroom and mutual interference can be avoided; high-frequency and low-frequency noise can be filtered out at the same time, only the sound of the user is returned, and sound distortion is not caused; and requirements of the wearing method and constituent materialsof the earphone are low; therefore, the effect of keeping quiet in a noisy environment can be obtained at any time, and the efficiency of oral learning is improved.

Description

technical field [0001] The invention relates to a sound-returning functional component, a sound-insulating feedback earphone, an application thereof, and a sound-insulating feedback method, belonging to the technical field of earphones. Background technique [0002] When multiple students practice speaking or reading at the same time in the same classroom, it will cause mutual interference and affect the learning effect. Wearing sound-isolating headphones or earplugs, there will be a sense of distortion. In the process of oral self-study, it is also very important to compare with the standard pronunciation and intelligently correct the pronunciation errors, but it is impossible for the teacher to pay attention to all the students all the time. Therefore, using smart earphones to solve such problems has certain practical significance. [0003] In the existing headphone noise reduction technology, there are mainly two solutions: active noise reduction and physical isolation....

AI Technical Summary

Problems solved by technology

However, with the deepening of research, the limitations of traditional speech recognition methods are gradually exposed, mainly difficult to solve the following problems:

[0006] (1) Environmental noise problem

When the signal-to-noise ratio decreases, the recognition error rate increases exponentially, and the current various anti-noise algorithms mainly rely on some filters in the frequency domain. Although the recognition performance under noise conditions has been improved to a certain extent, but The error rate increases under the condition of high signal-to-noise ratio

[0007] (2) Oral pronunciation problems

These phenomena lead to a double mismatch between the acoustic model and the language model, which leads to a sharp drop in the recognition rate.

[0008] (3) Multi-speaker problem

When multiple speakers speak at the same time, there is no good algorithm in the traditional algorithm that can separate the overlapping parts, and there is no speech recognition algorithm that supports simultaneous decoding of multiple targets.

Images