当前位置:网站首页>6.28 Dahua written examination
6.28 Dahua written examination
2022-07-27 02:12:00 【Yu Xiaohe】
Just finished the written examination of Dahua in advance in the morning , Make a note of .
Time of written examination 60 minute ,13 Problem ,10 Choose to fill in the blanks +3 A simple answer .
The question type is different from that of last year's written examination seen in Niuke , No programming problems .
Select the blank part
new delete And malloc free The difference between
x&(x-1) The calculation of
Correct initialization of two-dimensional array
Give the image input size of a certain layer of convolutional neural network , Use convolution kernel size , Step size , Calculate the output image size
Do convolution filtering on an image , Convolution kernels for 3*3 matrix , From left to right , From top to bottom 1,1,1,1,8,1,1,1,1 What is the operation ,-1,-1,-1,-1,8,-1,-1,-1,-1 What is the operation
rgb turn hsv,hsv Re turn rgb, How many degrees of rotation
The rest is basically to show the program what the output is
The short answer part
1. How to solve the problem that the complex model of deep learning algorithm is difficult to deploy in practical application
Give the data bit width , Input and output image size , Convolution kernel size , Calculate storage space and calculation amount
Change the data bit width 、 Image size 、 Convolution kernel type , How does the storage space and calculation amount change
2. Use the camera to shoot a scene containing moving objects 1ms Exposure map A and 16ms Exposure map B, Design an algorithm to synthesize two graphs into one
3. Two rectangles intersect , Give the coordinates of the upper left corner and the lower right corner , Calculate the coordinates of intersecting rectangles , And the best way to judge whether it intersects
边栏推荐
- [MySQL] MySQL startup and shutdown commands and some error reports to solve problems
- Docker advanced: detailed explanation of MySQL master-slave replication and redis cluster capacity expansion and reduction configuration cases
- Electron FAQ 61 - must the client run with administrator privileges?
- Mysql database - interview questions
- Unity Huatuo example project source code analysis and inspiration
- count(*)为什么很慢
- Beyond hidden display ellipsis
- STM32入门教程第二讲
- 详解文本生成图像的仿射变换模块(Affine Transformation)和条件批量标准化(CBN)
- MVCC及其原理详解
猜你喜欢

Difference between fat AP and thin AP & advantages and disadvantages of networking

科学计算库 —— Matplotlib

Unity Huatuo hot update environment installation and sample project

Text to image paper intensive reading rat-gan: recursive affine transformation for text to image synthesis

Hash索引和B+树相关知识

Text to image论文精读GR-GAN:逐步细化文本到图像生成 GRADUAL REFINEMENT TEXT-TO-IMAGE GENERATION

静态综合实验(静态路由、环回接口、缺省路由、空接口、浮动静态的综合练习)
![Unity Huatuo revolutionary hot update series [1]](/img/bc/ed480fc979c08c362784a3956d7fe2.jpg)
Unity Huatuo revolutionary hot update series [1]

2022年T2I文本生成图像 中文期刊论文速览-1(ECAGAN:基于通道注意力机制的文本生成图像方法+CAE-GAN:基于Transformer交叉注意力的文本生成图像技术)

Text to image intensive reading df-gan:a simple and effective baseline for text to image synthesis
随机推荐
count(*)为什么很慢
读写分离、主从同步一文搞定
7.13 蔚来提前批笔试
Flink1.13.6 detailed deployment method
7.8 锐捷网络笔试
初识网页设计
行,表,页,共享,排他,悲观,乐观,死锁
How can smart people leave without offending others?
测开基础 日常刷题 (持续更新ing...)
Specify that SQL only supports select syntax
The gradient descent method and Newton method are used to calculate the open radical
Vitgan: training Gans with vision transformers
Tinyint type map is received and returned as Boolean
详解文本生成图像的仿射变换模块(Affine Transformation)和条件批量标准化(CBN)
JS -- first understand the naming rules and data types of JS and variables
JS max?
三范式,约束,部分关键字区别,
6.30联发科笔试
HCIA基础知识(1)
OSPF的重发布及路由策略