funkidd的博客

PCM 转WAV。 2. 多通道pcm挑出其中

funkidd — Wed, 29 Dec 2021 14:50:32 +0800

import wave import os import wave import numpy as np def gen_wav(filepath): wavelist=[] filenames=os.listdir(filepath) for filename in filenames: print(filename) name,category=os.path.splitext(filepath+filename) #分解文件扩展名...

一个资料夹里面wav合并（同sample rate）

funkidd — Wed, 29 Dec 2021 14:47:21 +0800

import wave import os import wave import numpy as np filepath= "/Users/anker/Documents/2_Speech_data/anker设备实际录音/3382/Stage1/鼠标键盘/" wavelist=[] filenames=os.listdir(filepath) for filename in filenames[:]: name,category=os.path.splitext(filepath+file..

大音频档案拆小

funkidd — Wed, 29 Dec 2021 14:43:24 +0800

from pydub import AudioSegment import pydub from pydub.silence import split_on_silence import sys import os filename = '/Volumes/Speech/0DownlinkNoise/officenoise.wav' audio_segment = AudioSegment.from_file(filename, format='wav') total = int(audio_segmen

小档案经过RIR，归一化后，合成大档案

funkidd — Wed, 29 Dec 2021 14:41:41 +0800

import random import os import glob import numpy as np import scipy.io.wavfile as wav import soundfile as sf import librosa from scipy import signal import scipy.io as scio dataFile = 'rir.mat' rir = scio.loadmat(dataFile) def merge_files(path_read_folde

model.fit_generator

funkidd — Wed, 08 Sep 2021 09:15:18 +0800

def generator(): while 1: global cnt cnt += 1 print("cnt: ",cnt) x = X_train[(cnt-1)*batch_size:(cnt)*batch_size,:] y = Y_train[(cnt-1)*batch_size:(cnt)*batch_size,:] yield x,y def val(): while 1: ...

K-Fold

funkidd — Wed, 08 Sep 2021 09:14:14 +0800

https://blog.youkuaiyun.com/qq_22885109/article/details/81081977

VS code C 怎么编译、单步执行。。

funkidd — Mon, 15 Mar 2021 09:29:25 +0800

https://blog.youkuaiyun.com/u013163834/article/details/85218331

如何在github上下载单个文件夹？用git就可以呀

funkidd — Fri, 05 Mar 2021 17:06:36 +0800

https://www.jianshu.com/p/74a0441ed9b7

Mac m1 安装tensorflow

funkidd — Sun, 28 Feb 2021 09:40:17 +0800

最主要refernece这个链接。。。 https://blog.youkuaiyun.com/lyang1896/article/details/113187799 还有 https://www.ai-buzz.com/18-steps-to-install-tensorflow_macos-on-m1-macbook-2020/ 关键有个地方我一直 condaactivate tensorflow_env 失败：之前没有注意，后来发现了。。搞了超级久，要加这个：就可以过啦。 source...

m1 安装tensorflow

funkidd — Fri, 26 Feb 2021 16:41:11 +0800

https://blog.youkuaiyun.com/m0_46402436/article/details/111355602

2021-02-02 keras 可以改 loss（从 input 或者 tensor）

funkidd — Tue, 02 Feb 2021 18:38:58 +0800

from keras.callbacks import LearningRateScheduler from keras.layers import Multiply import keras from keras.models import Sequential from keras.models import Model from keras.layers import Input from keras.layers import Dense from keras.layers import LS.

Speex 介绍

funkidd — Tue, 12 Jan 2021 13:49:04 +0800

https://my.oschina.net/u/4340620/blog/3259325

rnnoise 两篇

funkidd — Wed, 06 Jan 2021 14:37:32 +0800

深度学习降噪方案-RNNoise源码解析 - 特征提取 https://blog.youkuaiyun.com/danteLiujie/article/details/102799038?utm_medium=distribute.wap_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-3.wap_blog_relevant_pic&depth_1-utm_source=distribute.wap_relevant.none-task-blog

makefile 基础

funkidd — Tue, 05 Jan 2021 15:35:57 +0800

# test:a.o b.o # gcc -o test a.o b.o # a.o : a.c # gcc -c -o a.o a.c # b.o : b.c # gcc -c -o b.o b.c test: a.o b.o c.o gcc -o test $^ %.o : %.c gcc -c -o $@ $< %.o：表示所用的.o文件 %.c：表示所有的.c文件 $@：表示目标 $<：表示第1个依赖文件 $^：表示所有依赖...

python 合并当前资料夹中所有wav档

funkidd — Thu, 31 Dec 2020 12:00:38 +0800

import wave import os import wave import numpy as np filepath="." wavelist=[] filenames=os.listdir(filepath) for filename in filenames: name,category=os.path.splitext(filepath+filename) #分解文件扩展名 if category=='.wav': #若文件为wav音频文件 wavelist...

58 直播针对 webrtc 跟 rnnoise 的研究/ webrtc 安卓移植

funkidd — Wed, 23 Dec 2020 14:51:39 +0800

https://kknews.cc/zh-hk/tech/o5k8rrq.html

Rnnoise 训练遇到的问题

funkidd — Wed, 23 Dec 2020 11:35:59 +0800

(1) cd src ; ./compile.sh (2) ./denoise_training clean.pcm noise.pcm 50000 > training.f32 (3) cd training ; python3 ./bin2hdf5.py ../src/training.f32 50000 87 training.h5 ###这个 50000跟上面是搭配的，可以调整training次数 (4) python3 rnn_train.py (5)python3 dump...

HRTF

funkidd — Thu, 10 Dec 2020 14:43:28 +0800

https://blog.youkuaiyun.com/qq_44494015/article/details/105076556?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522160758135019724838574396%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=160758135019724838574396&am

AEC————网站

funkidd — Thu, 10 Dec 2020 14:40:07 +0800

https://blog.youkuaiyun.com/ljl86400/article/details/80475781?utm_medium=distribute.pc_relevant.none-task-blog-title-2&spm=1001.2101.3001.4242 下面这个有代码 https://blog.youkuaiyun.com/weixin_44652758/article/details/107633117

语音增强---网站

funkidd — Wed, 16 Sep 2020 14:18:50 +0800

https://zhuanlan.zhihu.com/p/99143394