自定义博客皮肤VIP专享

*博客头图:

格式为PNG、JPG,宽度*高度大于1920*100像素,不超过2MB,主视觉建议放在右侧,请参照线上博客头图

请上传大于1920*100像素的图片!

博客底图:

图片格式为PNG、JPG,不超过1MB,可上下左右平铺至整个背景

栏目图:

图片格式为PNG、JPG,图片宽度*高度为300*38像素,不超过0.5MB

主标题颜色:

RGB颜色,例如:#AFAFAF

Hover:

RGB颜色,例如:#AFAFAF

副标题颜色:

RGB颜色,例如:#AFAFAF

自定义博客皮肤

-+

MachineLP的专栏

成功收获成果,失败收获智慧,投入收获快乐!

  • 博客(634)
  • 资源 (32)
  • 收藏
  • 关注

原创 MachineLP的CV与NLP

深度学习这些坑你都遇到过吗?(1)检测Faster RCNN:RPN,anchor,sliding windowsDL杂记:YOLOV3之禅(2)识别tf40:图像检索(triplet_loss)之Conditional Similarity Networks(3)OCROCR综述OCR工程git:https://github.com/MachineLP/OCR_repo1.tf20: CNN—识别字符验证码2.身份证识别——生成身份证号和汉字...

2020-05-17 11:07:03 1596

原创 MachineLP博客目录

MachineLP: 其实事物发展有自己的潮流和规律,当你身处潮流之中的时候,要紧紧抓住潮流的机会,想办法脱颖而出,即使没有成功,也会更加洞悉时代的脉搏,收获珍贵的知识和经验。而如果潮流已经退去,这个时候再去往这个方向上努力,只会收获迷茫与压抑,对时代、对自己都没有什么帮助。 但是时代的浪潮犹如海滩上的浪花,总是一浪接着一浪,只要你站在海边,身处这个行业之中,下...

2017-11-02 10:48:28 37118 9

原创 OpenCV之二值图像 联通组件寻找

python代码:import cv2 as cvimport numpy as npdef connected_components_demo(src): src = cv.GaussianBlur(src, (3, 3), 0) gray = cv.cvtColor(src, cv.COLOR_BGR2GRAY) ret, binary = cv.threshold(gray, 0, 255, cv.THRESH_BINARY | cv.THRESH_OTSU)

2021-08-11 15:04:03 205

原创 OpenCV之图像二值化与去噪

python代码:import cv2 as cvimport numpy as npdef method_1(image): gray = cv.cvtColor(image, cv.COLOR_BGR2GRAY) t, binary = cv.threshold(gray, 0, 255, cv.THRESH_BINARY | cv.THRESH_OTSU) return binarydef method_2(image): blurred = .

2021-08-11 14:59:18 876

原创 OpenCV之图像二化自适应阈值算法

python代码:import cv2 as cvimport numpy as np## THRESH_BINARY = 0# THRESH_BINARY_INV = 1# THRESH_TRUNC = 2# THRESH_TOZERO = 3# THRESH_TOZERO_INV = 4#src = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src

2021-08-11 14:56:29 231

原创 OpenCV之图像二值寻找算法 – TRIANGLE

python代码:import cv2 as cvimport numpy as np#import tensorflow as tf# tf.enable_eager_execution()## THRESH_BINARY = 0# THRESH_BINARY_INV = 1# THRESH_TRUNC = 2# THRESH_TOZERO = 3# THRESH_TOZERO_INV = 4#src = cv.imread("./test.png")cv.namedWin

2021-08-11 14:54:19 336

原创 OpenCV之图像二值寻找算法 – OTSU

python代码:import cv2 as cvimport numpy as np## THRESH_BINARY = 0# THRESH_BINARY_INV = 1# THRESH_TRUNC = 2# THRESH_TOZERO = 3# THRESH_TOZERO_INV = 4#src = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)

2021-08-11 14:52:31 116

原创 OpenCV之基本阈值操作

python代码:import cv2 as cvimport numpy as np## THRESH_BINARY = 0# THRESH_BINARY_INV = 1# THRESH_TRUNC = 2# THRESH_TOZERO = 3# THRESH_TOZERO_INV = 4#src = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)

2021-08-11 14:51:02 86

原创 OpenCV之二值图像介绍

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)T = 127# 转换为灰度图像gray = cv.cvtColor(src, cv.COLOR_BGR2GRAY)h, w = gray.shapeT = cv.mean(gray)[0]print("

2021-08-11 14:49:20 358

原创 OpenCV之图像模板匹配

python代码:import cv2 as cvimport numpy as npdef template_demo(): src = cv.imread("./test.png") tpl = cv.imread("./test01.png") cv.imshow("input", src) cv.imshow("tpl", tpl) th, tw = tpl.shape[:2] result = cv.matchTemplate(src,

2021-08-11 14:47:33 155

原创 OpenCV之拉普拉斯金字塔

python代码:import cv2 as cvimport numpy as npdef laplaian_demo(pyramid_images): level = len(pyramid_images) for i in range(level-1, -1, -1): if (i-1) < 0: h, w = src.shape[:2] expand = cv.pyrUp(pyramid_images

2021-08-11 14:46:04 191

原创 OpenCV之图像金字塔

python代码:import cv2 as cvdef pyramid_down(pyramid_images): level = len(pyramid_images) print("level = ",level) for i in range(level-1, -1, -1): expand = cv.pyrUp(pyramid_images[i]) cv.imshow("pyramid_down_"+str(i), expand)

2021-07-30 15:57:19 76

原创 OpenCV之Canny边缘检测器

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)# t1 = 100, t2 = 3*t1 = 300edge = cv.Canny(src, 100, 300)cv.imshow("mask image", edge)cv.imwrite("./edge.

2021-07-30 15:51:50 235

原创 OpenCV之USM 锐化增强算法

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)# sigma = 5、15、25blur_img = cv.GaussianBlur(src, (0, 0), 5)usm = cv.addWeighted(src, 1.5, blur_img, -0.5,

2021-07-30 15:50:43 305

原创 OpenCV之图像锐化

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)# sharpen_op = np.array([[-1, -1, -1], [-1, 9, -1], [-1, -1, -1]], dtype=np.float32)sharpen_op = np.array([

2021-07-23 15:31:06 449 1

原创 OpenCV之图像梯度 – 拉普拉斯算子(二阶导数算子)

python代码:import cv2 as cvimport numpy as npimage = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", image)h, w = image.shape[:2]src = cv.GaussianBlur(image, (0, 0), 1)dst = cv.Laplacian(src, cv.CV_32F, ksize=

2021-07-23 15:11:05 524

原创 OpenCV之更多梯度算子

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)robert_x = np.array([[1, 0],[0, -1]], dtype=np.float32)robert_y = np.array([[0, -1],[1, 0]], dtype=np.float

2021-07-21 20:18:32 69

原创 OpenCV之Sobel算子

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)h, w = src.shape[:2]x_grad = cv.Sobel(src, cv.CV_32F, 1, 0)y_grad = cv.Sobel(src, cv.CV_32F, 0, 1)x_grad

2021-07-21 20:10:41 164

原创 OpenCV之自定义滤波器

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)blur_op = np.ones([5, 5], dtype=np.float32)/25.shape_op = np.array([[0, -1, 0], [-1, 5,

2021-07-21 20:06:07 263 1

原创 OpenCV之快速的图像边缘滤波算法

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)h, w = src.shape[:2]dst = cv.edgePreservingFilter(src, sigma_s=100, sigma_r=0.4, flags=cv.RECURS_FILTER)re

2021-07-21 20:04:05 240 1

原创 OpenCV之图像积分图算法

python代码:import cv2 as cvimport numpy as npdef get_block_sum(ii, x1, y1, x2, y2, index): tl = ii[y1, x1][index] tr = ii[y2, x1][index] bl = ii[y1, x2][index] br = ii[y2, x2][index] s = (br - bl - tr + tl) return sdef blur_

2021-07-21 19:53:27 116

原创 [语音识别] 文本加标点--BERT预测

文本加标点--训练BLSTMbert4keras==0.5.9# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippets

2021-07-19 20:07:29 648 2

原创 [语音识别] 文本加标点--训练BERT

文本加标点--训练BLSTMbert4keras==0.5.9# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippe.

2021-07-19 20:04:29 209

原创 [语音识别] 文本加标点--LSTM预测

文本加标点--训练BLSTMbert4keras==0.5.9# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippets

2021-07-19 20:00:23 387 1

原创 [语音识别] 文本加标点--训练BLSTM+ATTENTION

文本加标点--训练BLSTMbert4keras==0.5.9# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippets

2021-07-19 19:51:20 168

原创 [语音识别] 文本加标点--训练BLSTM

文本加标点--训练BLSTMbert4keras==0.5.9# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippets

2021-07-19 19:49:19 120

原创 [语音识别] 文本加标点--训练LSTM

文本加标点--训练LSTMkeras版本:# -*- coding:utf-8 -*-import os import reimport sysimport timeimport jiebaimport gensim import loggingimport numpy as npimport pandas as pdfrom tqdm import tqdmimport multiprocessingfrom bert4keras.snippets import ..

2021-07-19 19:46:16 224 1

原创 [语音识别] 文本加标点--准备训练数据 (适用于LSTM与BERT)

Paper: U2: Unified Streaming and Non-streaming Two-pass End-to-end Model for Speech Recognition WeNet: Production First and Production Ready End-to-End Speech Recognition Toolkit, v1. WeNet: Production Oriented Streaming and Non-streaming En.

2021-07-19 19:19:19 312 2

原创 OpenCV之边缘保留滤波算法 – 均值迁移模糊(mean-shift blur)

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)h, w = src.shape[:2]dst = cv.pyrMeanShiftFiltering(src, 15, 30, termcrit=(cv.TERM_CRITERIA_MAX_ITER+cv.TERM

2021-07-18 15:06:29 133

原创 OpenCV之边缘保留滤波算法 – 高斯双边模糊

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)h, w = src.shape[:2]dst = cv.bilateralFilter(src, 0, 100, 10)result = np.zeros([h, w*2, 3], dtype=src.dtyp

2021-07-18 15:03:03 74

原创 OpenCV之图像去噪声

python代码:import cv2 as cvimport cv2 as cvimport numpy as npdef add_salt_pepper_noise(image): h, w = image.shape[:2] nums = 10000 rows = np.random.randint(0, h, nums, dtype=np.int) cols = np.random.randint(0, w, nums, dtype=np.int)

2021-07-18 14:53:00 732

原创 OpenCV之图像噪声

python 代码:import cv2 as cvimport numpy as npdef add_salt_pepper_noise(image): h, w = image.shape[:2] nums = 10000 rows = np.random.randint(0, h, nums, dtype=np.int) cols = np.random.randint(0, w, nums, dtype=np.int) for i in rang

2021-07-18 14:36:00 268

原创 OpenCV之中值模糊

import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)dst = cv.medianBlur(src, 5)cv.imshow("blur ksize=5", dst)cv.waitKey(0)cv.destroyAllWindows()#include <open.

2021-07-18 14:32:05 76

原创 OpenCV之图像均值与高斯模糊

python代码:import cv2 as cvimport numpy as npsrc = cv.imread("./test.png")cv.namedWindow("input", cv.WINDOW_AUTOSIZE)cv.imshow("input", src)dst1 = cv.blur(src, (5, 5))dst2 = cv.GaussianBlur(src, (5, 5), sigmaX=15)dst3 = cv.GaussianBlur(src, (0,

2021-07-18 14:30:45 65

原创 OpenCV之图像卷积操作

python代码:import cv2 as cvimport numpy as npdef custom_blur(src): h, w, ch = src.shape print("h , w, ch", h, w, ch) result = np.copy(src) for row in range(1, h-1, 1): for col in range(1, w-1, 1): v1 = np.int32(src[

2021-07-18 14:29:56 211

原创 OpenCV之图像直方图反向投影

python代码:import cv2 as cvimport numpy as npfrom matplotlib import pyplot as pltdef back_projection_demo(): sample = cv.imread("./Mat.png") # hist2d_demo(sample) target = cv.imread("./test.png") # hist2d_demo(target) roi_hsv = cv

2021-07-18 14:23:58 76

原创 [语音识别] kaldi -- aidatatang_200zh脚本解析:三音速详解

# train tri1 [first triphone pass]steps/train_deltas.sh --cmd "$train_cmd" \2500 20000 data/train data/lang exp/mono_ali exp/tri1 || exit 1;# steps/train_deltas.sh <num-leaves> <tot-gauss> <data-dir> <lang- dir> <alignment-di.

2021-07-17 00:30:12 228

原创 [语音识别] kaldi -- aidatatang_200zh脚本解析:获取对齐文件(对指定的数据进行对齐,作为新模型的输入)

输入:tree & final.mdl & text & L.fst & feats输出:ali.JOB.gz (根据重新构建的图产生) & final.mdl (cp from 训练过程的最终结果)steps/align_si.sh --cmd "$train_cmd" --nj 10 \ data/train data/lang exp/mono exp/mono_ali || exit 1;流程:1. 根据$use_graphs(默认是

2021-07-17 00:22:58 204

原创 [语音识别] kaldi -- aidatatang_200zh脚本解析:单音素解码

输入:feats & final.mdl & HCLG.fst输出:lat.JOB.gz## 在dev数据集上执行steps/decode.sh --cmd "$decode_cmd" --config conf/decode.config --nj 10 \ exp/mono/graph data/dev exp/mono/decode_dev## 在test数据集上执行steps/decode.sh --cmd "$decode_cmd" --config conf/

2021-07-17 00:08:54 189

原创 [语音识别] kaldi -- aidatatang_200zh脚本解析:构建解码图

输入:final.mdl & tree & L_disambig.fst & G.fst输出:HCLG.fst (在exp/xxxx/graph下)## 构建解码图### mkgraph.sh主要生成了HCLG.fst,后续识别主要利用了三个文件,分别是final.mdl、HCLG.fst、words.txt(从lang文件夹 cp过来)。utils/mkgraph.sh data/lang_test exp/mono exp/mono/graph || exit 1

2021-07-17 00:05:51 196

训练语言模型的样本

训练语言模型的样本文件,训练语言模型的样本文件,训练语言模型的样本文件,训练语言模型的样本文件。

2018-06-03

训练机器学习模型代码

训练机器学习模型代码,训练机器学习模型代码,训练机器学习模型代码

2017-12-18

Make成python可以调用的.so文件。

Make成python可以调用的.so文件,并且包含例子。Make成python可以调用的.so文件,并且包含例子。

2017-11-30

opencv2.4.8 源码

opencv2.4.8 源码,opencv2.4.8 源码,opencv2.4.8 源码。

2017-11-29

黄文坚,唐源 TensorFlow实战代码

TensorFlow实战代码 TensorFlow实战代码 TensorFlow实战代码

2017-10-18

Face Recognition(face_recognition) Using Hadoop Streaming API

Face Recognition(face_recognition) Using Hadoop Streaming API Face Recognition(face_recognition) Using Hadoop Streaming API

2017-10-08

Face_Detection_Using_Hadoop

Face_Detection_Using_Hadoop Face_Detection_Using_Hadoop

2017-10-08

matlab使用TCP/IP Server Sockets

matlab使用TCP/IP Server Sockets

2017-08-31

RNN生成古诗词

RNN生成古诗词

2017-03-15

简单的聊天机器人

简单的聊天机器人。

2017-02-26

深度学习破解字符验证码

深度学习破解字符验证码。

2017-02-23

叶子数据leaf_data

leaf_data

2017-02-10

Linux命令速查手册

Linux命令速查手册

2017-01-24

是AI就躲个飞机-纯Python实现人工智能下载

是AI就躲个飞机-纯Python实现人工智能

2017-01-17

Python简明教程-沈洁元

Python简明教程

2017-01-04

caffe深度学习薛开宇笔记实例-基于卷积神经网络的声音识别

目前的音乐检索系统用流派、风格、情感等类别标签检索音乐。其中,如果人工标注 音乐这些类别标签,则存在主动性强、费时费力、速度慢的问题,而如果采用传统的自动 标注方式,则存在准确率低的问题。后者准确率低的原因是,其标注时使用的模型不能 很好识别音乐。随着 Hinton 提出深度学习模型后,因其在图像和语音识别领域均取得很 好的成果,在识别领域成为了研究热点。因此,本文旨在研究如何使用深度学习中的卷 积神经网络(Convolutional Neural Networks,CNN)模型,设计出一个准确度高、速度快 的自动音乐分类系统,用作标注音乐的类别标签。

2016-12-27

Deep Learning textbook

The Deep Learning textbook is a resource intended to help students and practitioners enter the field of machine learning in general and deep learning in particular.

2016-12-09

CNN实现手写数字库mnist

CNN实现手写数字库mnist

2016-11-29

opencv_traincascade训练分类器,手势识别

opencv_traincascade训练分类器,手势识别

2016-11-27

openCV实现目标检测

openCV实现目标检测

2016-11-27

文件中的图片批量处理,设置想要的尺寸大小。

文件中的图片批量处理,设置想要的尺寸大小。

2016-11-27

基于粒子滤波的物体跟踪

基于粒子滤波的物体跟踪

2016-11-25

Mac+Xcode+openCV实现OpenTLD

OpenTLD(Tracking-Learning-Detection)运动跟踪与检测开源软件

2016-11-25

迅速上手,能随便跑起来的OpenTLD

OpenTLD(Tracking-Learning-Detection)运动跟踪与检测开源软件。

2016-11-25

训练xml文件时,产生info.txt文件

用于生成正样本描述的objectmarker(附源码) 鼠标选定区域,空格记录,回车退出程序并保存。 注意:每次保存都以info.txt命名,所以运行前的info.txt会被重写。

2016-11-01

tensorflow 学习资料

为了进一步激发你的学习欲望, 我们想让你先看一下TensorFlow 是如何解决一个经典的机器学习问题的.在神经网络领域,最为经典的问题莫过于MNIST 手写数字分类.为此,我们准备了两篇不同的教程, 分别面向初学者和专家.如果你已经使用其它软件训练过许多MNIST 模型,请参阅高级教程(红色药丸).如果你以前从未听说过MNIST,请先阅读初级教程(蓝色药丸).如果你的水平介于这两类人之间, 我们建议你先快速浏览初级教程,然后再阅读高级教程.

2016-10-18

斯坦福大学机器学习课程个人笔记完整版

2016-04-19

扣人脸和人脸标定的matlab代码

扣人脸和人脸标定的matlab代码

2016-04-19

kinect v2获取高清面部帧

kinect v2获取高清面部帧

2016-04-07

kinect v2 语音获取

kinect v2 语音获取

2016-04-07

openCV人脸检测的C++代码

openCV人脸检测的C++代码

2016-04-07

读取文件中所有图片路径的完整C++代码。

读取文件中所有图片路径的完整C++代码。

2016-04-07

空空如也

TA创建的收藏夹 TA关注的收藏夹

TA关注的人

提示
确定要删除当前文章?
取消 删除