QA Official

A standard resnet network with 18-150 layers 2019-04-03
import torch.nn as nn import torch.utils.model_zoo as model_zoo __all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101', 'resnet152'] model_urls = { 'resnet18': '', 'resnet34': '', 'resnet50': '', 'resnet101': '', 'resnet152': '', } def conv3x3(in_planes, out_planes, stride=1): """3x3 convolution with padding""" return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride, padding=1, bias=False) def conv1x1(in_planes, out_planes, stride=1): """1x1 convolution""" return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False) class BasicBlock(nn.Module): expansion = 1 def __init__(self, inplanes, planes, stride=1, downsample=None): super(BasicBlock, self).

FasterRCNN Configuration Process under Windows 7 64 Bit 2019-04-03
First of all, one of the most important prerequisites is to have the VC++operating environment and components in advance. It is best to install vs2012 or later in advance. After having the VC++operating environment and components, try the following steps to configure fast RCNN. If you are not sure whether there is a VC++operating environment that meets the requirements, you can try it first. If you fail and worry about the next configuration, you can directly uninstall Adacconda and reinstall Adacconda and then re-install it according to the procedure.

Fine tuning training is carried out by using Oxford trained face model;Identify 3 faces 2019-04-03
problems encountered: Target: Fine tuning training is carried out by using Oxford trained face model;Identify 3 faces deep learning framework: Caffe network model: VGG16 Current situation: the collected data is not enough, the network can run, but the effect is not good, the next step may be to increase the amount of data;Learning to adjust parameters Work Flow # Python代码参照

Java NIO Framework Netty Tutorial (1)-Hello Netty 2019-04-03
Let's start with a few words. If you don't know what Netty does or can do.That can be a simple search to understand first.I can only say Netty is a NIO framework that can be used to develop distributed Java programs.You can try your best to imagine what you can do.Technology serves people, not limits them. If you already have everything, let's start with a piece of code.The first step

Netty Beginner-Level 1 (Basic Operation) 2019-04-03
Netty Getting Started 1. Why netty? Netty is the implementation framework based on java NIO, which is exactly implemented based on NIO2.0, that is, AIO. However, because the use of NIO is complicated, netty appears. The NIO server and client developed through netty are very simple, with only a few tens of lines of code, NIO functions that can only be realized by hundreds of lines of code can be

Netty Game Server Actual Combat Development (2): Bytebuf-Container of Byte Data 2019-04-03
buffer The basic unit of network data is always byte.Java NIO provides ByteBuffer as byte container, but its function is too limited and has not been optimized.Using ByteBuffer is usually a tedious and complicated matter. Fortunately, Netty provides a powerful buffer implementation class to represent byte sequences and to help you manipulate bytes and custom POJO.This new buffer class, ByteBuf, is as efficient as JDK's ByteBuffer.ByteBuf was designed to

Netty Getting Started Tutorial (2)-Netty Server and Client 2019-04-03
Netty version in the code is 3.x because it is a preliminary study of Netty and Netty3 is easier to understand from the information found on the internet. 1. Netty Server import; import java.util.concurrent.ExecutorService; import java.util.concurrent.Executors; import org.jboss.netty.bootstrap.ServerBootstrap; import; import; import; import; import org.jboss.netty.handler.codec.string.StringDecoder; import org.jboss.netty.handler.codec.string.StringEncoder; /** * netty服务端入门 * */ public class Server { public

Netty from Zero to One-Getting Started 2019-04-03
Start This chapter explains the core components of Netty in turn and includes simple examples to get you started quickly.At the end of this chapter, you can write the client and server that depend on Netty. preparation If you want to run the examples in this chapter, you need at least the latest version of Netty and JDK above 1.6.Please download the latest Netty to, and JDK please download

Python Third Party Library-MatPlotlib Library 2019-04-03
import matplotlib.pyplot as plt pyplot module plt.ion (): Opens Interactive Mode plt.figure(num=None, figsize=None, dpi=None, facecolor=None, edgecolor=None, frameon=True, FigureClass=<class 'matplotlib.figure.Figure'>, **kwargs) Parameters: Num: integer or string, the default value is None.Id of the figure object.If num is not specified, a new figure will be created, and the id (that is, the number) will be incremented. This id exists in the member variable NUMBER of the figure object.If a num value is

Pytorch-based Feature Map Extraction 2019-04-03
brief introduction In order to understand the operation process of convolutional neural network conveniently, it is necessary to visually display the operation results of convolutional neural network. can be roughly divided into the following steps: Extraction of Single Picture Neural Network Construction Feature Map Extraction Visual Display Extraction of Single Picture According to the requirements of the target, convolution operation is required for a single picture, but the data read