Zihao 2021-01-13
2021-01-18 23:17
[TOC]
My name is Zihao, a student from Tongji University, Shanghai.
I am both a Most Valuable Professional(MVP) and Microsoft Student Ambassador(MSA) in China.
I am also a student leader in Microsoft Student Club at Tongji University.
And I was also an intern in Microsoft Research Asia.
I have a video channel about AI, Python, Machine Learning, Deep Learning, Computer Vision, and raspberry pi. Also have many tutorials of Azure and Cognitive Service such as Custom Vision.
In this special period of time, thanks for giving me this precious opportunity.
I am looking forward to sharing my experience with technology content creation and the open source community.
That's a really interesting story.
I majored in civil engineering when I was an undergraduate.
To be honest, at that time, I hate coding, I never write a single line of code.
I thought coding is exhausting, especially debugging is something hard and boring.
My major, civil engineering, focuses on all kinds of construction, design and build architecture, house, tunnel, bridge, railway, tower, nothing but mechanics and dynamics.
There have already been lots of ready-made technical standards or software, you just follow the rules and nothing special, nothing new.
This is my graduation project, I spent 3 months designing an office building.
The thing changed in 2018 when I was an exchange student at the University of Pennsylvania,
There was a lecture in Wharton school, inviting the CEO of Microsoft, Nadella
He shared his insight about blockchain, quantum computing as well as open-source.
Later, I find Github, this amazing website.
I learned several open-source toolkits such as vscode, Keras, even bitcoin.
In Github, you can cooperate with developers all over the world.
Programming really can change the world, in my hands.
Just like, I can be the god of my own universe.
I started to study Python, data analysis, algorithms, machine learning, AI, deep learning, computer vision, front-end development with hunger.
That was not an easy way, I used to get stuck in a python bug or deal with data structure issue the whole day.
Without eating, sleeping, even drinking water, that's called heart-stream, isn't it?
And I started to write technical blogs.
At the same time, more and more people, especially students from traditional majors like me, such as mechanics, finance, medical research, energy engineering, automobile, want to get into the IT world, too.
Use advanced machine learning and AI to empower their own profession.
With my experience, I believe I can really help them.
simplify the learning curve, avoid detours, and consequently save lots of time and money as well.
My first video is about convolutional neural networks, the breakthrough algorithm in computer vision, and deep learning.
Later, I find myself good at teaching this kind of hard-core thing.
More and more subscribers make me more confident and I think this could be the career of my lifetime.
We aim to help zero basic beginners and cross-fields- learners quickly master AI and data mining in easy-to-understand-way, and quickly apply what they have learned
Consequently, provide a complete system of AI education solutions for students, job seekers, employees, and large enterprises. Help beginners to cross the threshold of mathematics, algorithm programming, frontier exploration, and application landing, quickly apply what they have learned.
My video channel covers all kinds of fields related to computer science.
from math, machine learning basics, deep learning, computer vision, python programming, data analysis, data visualization, edge computing.
Especially we focus on computer vision.
Computer vision is telling a computer how to extract or recognize information from images or videos.
Computer vision is a basic application in AI.
It contains image classification, object detection, semantic segmentation, instance segmentation, face recognition.
Keypoint detection, object tracking, 3D vision. and computer vision is widely used in different scenarios such as self-driving cars.
This is object detection, drawing rectangle boxes, based on the YOLOV5 algorithm.
I am also one of the contributors to this repository on Github.
There are four crucial steps for an AI developer:
- Math(linear algebra, probability statistics, Calculus)
- Coding and programming(python,scikit-learn,Keras)
- Algorithm(MLP,CNN,RNN,GAN)
- Real project(get your hands dirty)
Actually, if you already have experience in front-end development or robotics.
Or you have already been an expert in your own field.
Learning programming and machine learning is definitely a worthy investment.
For example, if you are an agriculture expert. This is a traditional job with a long history.
You can utilize new technology such as UAV drones, remote sensing, and computer vision to do lots of things, breaking the ceiling of traditional agriculture.
To be honest, you even don't need to write code and learn algorithms. There are lots of out-of-box toolkits such as custom vision in Azure Cognitive Service. You can just upload your datasets, train, and deploy the AI models just by clicking the mouse.
I majored in civil engineering, after learning machine learning, I can utilize data mining to empower this traditional profession.
As we know, China has a huge high-speed railway and urban subway system.
How do we detect and maintain the health of infrastructure and track?
That's what I can do to break the ceiling of traditional engineering.
Computer vision and object detection can give us a data-driven solution of detecting damage, cracking, shelling, checking, corrugation, fatigue fracture.
We can expect that in the future, everything is related to data.
This is exactly the key point that I want to share with you today.
No matter what your technical background is, no matter what domain or profession do you major in.
Interdisciplinary with machine learning and data mining will produce huge value in the future.
This could be your personal core competitiveness.
In addition, if you want to step into the academic palace of AI,
you need to read lots of paper from the top academic conferences
I have packed classical papers in AI, especially computer vision,
I took detailed notes of every formula, every graph, every referenced paper.
Also extended reading material.
So if you are a mater of PhD,
It is a great material which could save you lots of time.
Cognitive Service
Custom Vision
This process is extremely enjoyable for us. It is determined from the bottom of my heart that we are doing a great thing and providing high-quality educational resources for a huge number of future learners. No matter from national policies, public science popularization, IP media brands, knowledge creation, also making money, of course, it will surely create huge long-term value.
Open-source community and technology will thoroughly reshape the form of cooperation and knowledge production in the future, and consequently, change everyone's life.
Especially in China, open-source became more and more popular, developers use Github.
Not only do they use an open source developing toolkit such as visual studio code.
But also contribute their power to the community.
For instance.
At the early age of the covid-19 pandemic, middle January of 2020. Precisely the second day Wuhan this city was locked down.
My PhD mentor Wang Wei and my friend Zhao Shengyu, set up an open source project called Wuhan2020. They put their white paper and code on Github.
This project gathers different information about hospitals, hotels, factories, donation as well as mask and medical resource. Arrange that precious resource with high efficiency and transparency.
Hundreds of volunteers took part in this project and built 23 repositories on Github.
https://community.wuhan2020.org.cn/en-us/
https://github.com/wuhan2020/wuhan2020
I also contribute in dealing with this pandemic.
We develop an AI project, helping doctors to diagnose child pneumonia.
With the data from Kaggle and computation power from Custom Vision, we make this Web demo.
Doctors can upload X-ray images and AI will judge whether have pneumonia caused by a virus, bacteria, or health.
The world is rapidly changing,
Facing the sudden coronavirus pandemic,
After a year of hardship, we can understand more than ever the significance of a community with a shared future for mankind.
Not only the community of open-source or technology developers.
But also the community of common destiny for all mankind.
Microsoft supports Microsoft student club in hundreds of universities in China, also in my college, Tongji Univerisity.
We have the opportunity to work with Microsoft experts from MSRA on various cutting-edge projects.
Also many technical lectures, hackathons,s and summer camps.
Working in Microsoft is a very happy thing
I was an intern in ARD-Incubation Group,
We make a whole open-source community of AI education, called AI-EDU
from basics math to machine learning, neural networks, deep learning
https://github.com/microsoft/ai-edu
MVP means most valuable professional
MVP is a technical expert and or community leader, as well as open-source advocators.
Not only do they write blogs, make videos, but also they will give feedback on Microsoft Product.
There are several MVP sub-fields you can choose to apply.
Including Cloud, Azure, AI, .NET, SharePoint, Skype, Office 365, Windows
We also take part in the important activity in the community such as
Microsoft Ignite Conference
Microsoft Build Connect conference
AI Bootcamp
MSA means Microsoft Student Ambassadors
MSA members come from university students all over the world
Some of my Representative work videos, such as :
- CS231N computer vision Chinese lecture(https://www.bilibili.com/video/BV1K7411W7So)
- AI classic paper reading(https://www.bilibili.com/video/BV1r7411X7LC)
- Kaggle data mining(https://www.bilibili.com/video/BV1Wf4y1U7EL)
- Python basic(https://www.bilibili.com/video/BV1P5411W7g5)
- Raspberry Pi hardware basic(https://www.bilibili.com/video/BV1ab411x7aC)
- Simple tutorials of Microsoft cognitive service(https://www.bilibili.com/video/BV1CE411e7XR)
has gained great reputation in AI community.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent