No self-introduction
Loading
Profile
Follow
codes (2)
Speech enhancement algorithm based on minimum ener
no vote
Application background Speech is an important carrier of human information communication, but in many voice communication systems will inevitably be subject to noise, including noise and communication transmission equipment internal power noise. Severe noise interference can greatly reduce the intelligibility and intelligibility of speech, resulting in the decline of the quality of communication. Speech enhancement is an important technology in speech processing, its main purpose is from the noisy speech extract pure speech, provided high signal-to-noise ratio, and is a kind of speech signal to noise and interference of effective means to reduce the. Speech enhancement algorithm based on short time spectrum estimation is a common method in speech enhancement. The key point is the estimation of the background noise power spectrum. The existing methods are not good. The subject of this thesis is to find a more effective method of speech denoising based on the analysis
28164351 0 0
1
Scene matching algorithm for digital hearing aids
no vote
Application background Auditory scene matching algorithm is an important research topic in intelligent digital hearing aids. At present, all the algorithms are ideal, but the complexity of the algorithm is generally high and the experimental scene is not much, and is not suitable for real-time and practical requirements of the hearing aid algorithm. In this paper, we use the endpoint detection algorithm to reduce the computational complexity, which is based on the endpoint detection algorithm for speech and non speech segments, and to use the weighted minimum distance classifier with low computational complexity.
28164351 1 0
1
No more~
Follows0
Fans0