第一周小白的独白

ART.


Photo via The Paper A Chinese AI company affiliated with the Chinese Academy of Sciences on Tuesday announced a new breakthrough in 3D human reconstruction technology, bringing down surface error by 30%, which is expected to make the advanced technology available on smartphones. The company, Guangzhou-based Cloudwalk, announced that their breakthrough has set a new world record on 3D human reconstruction technology to build a single-frame image. On the key indexes of Human3.6M, Surreal, and UP-3D, the company decreased the surface error – a main indicator to measure the algorithm – at the millimeter level. On Human3.6M, for example, the surface error of human joints has dropped from 59.9mm to 46.7mm. In addition, execution speed has also been cut from 100 milliseconds to just 5 milliseconds, the company added. Wu Fan, a scientist with the company, said they are using a brand new 3D information representation model that can predict human body postures by analyzing the three primary color images and can depict the human body with over 60,000 points, China Science Daily reported. The breakthrough has reduced the requirement for input images. The technology can utilize ordinary optical cameras as sensing devices, which means applications that used to face restrictions in implementation can be gradually implemented in full and related smart image applications will be promoted, according to the company.

通过纸张拍照 周二,中国科学院下属的一家中国人工智能公司宣布了3D人体重建技术的一项新突破,将表面误差降低了30%,预计这项技术将在智能手机上应用。 这家总部位于广州的公司Cloudwalk宣布,他们的突破创造了一项3D人体重建技术的新世界纪录,用于构建单帧图像。在Human3.6米、超现实和up-3d的关键指标上,该公司降低了毫米级的表面误差——测量算法的主要指标。 以人体3.6米为例,人体关节表面误差由59.9毫米下降到46.7毫米。 此外,该公司还补充说,执行速度也从100毫秒降低到了5毫秒。 据《中国科学日报》报道,该公司科学家吴凡(音译)表示,他们正在使用一种全新的三维信息表示模型,通过分析三幅原色图像,可以预测人体姿势,并能以超过6万点的速度描绘人体。 这一突破降低了对输入图像的要求。该技术可以利用普通光学相机作为传感设备,这意味着在实施过程中面临限制的应用可以逐步全面实施,并将推广相关的智能图像应用。

The passage comes from:http://en.people.cn/n3/2019/0322/c90000-9559423.html\

https://blog.csdn.net/mrlixuec/article/details/85334193这个网址简单介绍了我们学习c++这门语言的流程,很明显我们前几周学习的仅仅是c++这门语言的第一部分<其基本数据类型和表达式>所以对于我而言课外学习,自主探索和交流讨论是很有必要的。

A文章编程实现两个正整数的除法,不能用除法操作符。

#include

 int div(const int x, const int y)

 { int dividend = x, multi, result = 0; while(dividend >= y)

 { multi = 1; while( multi * y <= (dividend >> 1) ) { multi <<= 1; } result += multi; dividend -= multi * y;

    } return result;

 }

int main()

{ int x, y; while( scanf("%d %d", &x, &y) != EOF ) { printf("%d÷%d=%d/n", x, y, div(x, y)); } return 0; };

div被除数 multi乘数

c程序是通可以过将数转化为字符串实现两数之间的除法。虽然这个程序没有直接使用除法运算符来的比较方便,但是更为客观的反映了程序解决问题的多样性。

A:char数组给char数组赋值不能直接用=必须用循环各位赋值。

B:对于vector容器来说,可以通过三种方法将一个vector容器的值赋给另一个vector容器,分别是

 1、通过等号直接赋值;

 2、通过assign函数赋值;

3、通过for循环进行赋值;

这几周看了20页相关的程序书籍逐渐的对这门任重而道远的科学产生了一点兴趣。同时也认识一些大佬互相讨论,相互学习,非常有心耐心讲解。作为一个刚上道德计算机小白欠缺的知识很多,需要从很基础开始理解,希望通过自己对程序慢慢的理解能够逐渐跟上大佬们的脚印。

你可能感兴趣的:(第一周小白的独白)