介绍：

本仓库包含利用反射声来实现语音增强的代码实现

20200107 The current version read data from storage. Network input: 5 beams label: The corresponding direct beam of 32 mic (already aligned) This way is somewhat ill.

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
log		log
models		models
pictures		pictures
DataPreprocessor.py		DataPreprocessor.py
README.md		README.md
abs_spec_mean.npy		abs_spec_mean.npy
abs_spec_std.npy		abs_spec_std.npy
calculate_cmvn.py		calculate_cmvn.py
config.py		config.py
const.py		const.py
encode.py		encode.py
enhance.py		enhance.py
evaluation.py		evaluation.py
handler.py		handler.py
hoa_params.py		hoa_params.py
net.py		net.py
norm.sh		norm.sh
signalprocess.py		signalprocess.py
test.py		test.py
test_get_az.py		test_get_az.py
test_time.py		test_time.py
train.py		train.py

cjf1699/Speech-Enhancement-Using-Early-Reflections