With these features that Meng Fanqi added today, Baidu is basically ready to cope with the first wave of government demand.

The function entrusted by Robin Li can be regarded as complete.

Indeed, as Wang Kai worried, if Baidu continues to focus on the project and the technical precipitation is not enough, it will still encounter the same problem in the follow-up needs.

It's just that this matter is a bad thing for Baidu, but not for Meng Fanqi.

Of course, whiteness is not completely devoid of technical strength, for example, on the basis of the vehicle being detected, the whiteness of functions such as recognizing license plates can be done.

At most, Meng Fanqi is a lot worse in speed and accuracy, but at this stage, others do not have Meng Fanqi's technical level, which is actually completely acceptable.

"If the Skynet project is going well, the next step should be face detection, right?" Meng Fanqi began to anticipate the needs of the government.

AI has gained momentum since then, but there are not so many applications in government projects.

Among them, the largest, most widely used, and the core of the demand should be face recognition.

This technology has been concerned by the government for a long time, and has been used to prevent and combat major criminal offenses since 01, and has been put into use on a large scale in 08.

However, these methods are all products of the past based on traditional methods.

After Alix unveiled the era of deep learning, Meng Fanqi led the real deep network, and the speed and accuracy of face recognition can have a qualitative leap.

According to the original timeline, in June 2014, SenseTime independently developed the DeepID series of face recognition algorithms, with an accuracy rate of 98.52.

While beating Facebook's DeepFace in the same period, DeepID's face recognition capability surpassed the human eye recognition level for the first time, which also broke through the warning line of large-scale industrialization.

In August of the same year, SenseTime received tens of millions of dollars in financing, and in October, SenseTime Technology was officially established.

By 2023, SenseTime's market value will be close to 100 billion, and it will initially rely on facial recognition technology that surpasses the human level.

That timeline, the residual network has not yet been proposed, Meng Fanqi thought for a while, at this time to make his own move, even if he doesn't use too much later technology, with the residual network + ArcFace, it should not be difficult to achieve 99.5+ recognition ability.

Is it necessary to start a short-term business? Meng Fanqi hesitated in this matter, and at this time, the technical strength must be enough to win the face intelligence part of the government system.

It is not clear whether the rash creation of a small company with a small number of people is qualified to have access to the person in charge of the government or the department in charge.

If you are not qualified to enter the game, the company will be a little weak, and in the end it may be the result of being acquired.

emmmm, if you think about it, it seems acceptable? Meng Fanqi looked at Wang Kai thoughtfully.

A very key thing is that many AI technologies cannot be monetized in time after they are made, such as the basic technology of ChatGPT, the attention mechanism - > Transformer - > GPT - > GPT2 - > GPT3 - > GPT3.5 (ChatGPT) - >GPT4.

The technical upgrade along the way, the first five versions are pure money-burning, and you may not be able to afford it.

But in a big company like Google, such results are valuable.

In order to lay out, many technologies he had to develop first, and in this situation, it is a better decision to wait for the time to be ripe in the large factory.

However, the huge Skynet project and the funds that arrived far earlier than expected made Meng Fanqi change his plans.

Meng Fanqi thought that he would only be noticed at the end of 13, and it would not be until the middle and late 14 years to accumulate tens of millions of wealth, but he never thought that 13 years were not over, and he had accumulated a lot of funds in hand and was qualified to participate.

Meng Fanqi focuses too much on the shocking technology after 2020, and ignores the ability of technology such as face recognition to land in the early stage.

If he can take over the face algorithm part of Skynet, his products will be all over China, and the company will also become an officially endorsed technology company.

After explaining the code part of the whiteness multi-vehicle detection and tracking, Meng Fanqi called Wang Kai to the corner of the office, leaning against the window.

Wang Kai, a muscular man of more than 1.8 meters, looks like he is blocked at the foot of the wall by Meng Fanqi, who is about 175, at the moment, and the picture is a little weird.

"I thought about it carefully, and Skynet's demand for face technology is a good opportunity that is hard to come by. It's also a share with a big factory, and now it's not bad for money, and some things are better than taking them yourself. ”

The former has a smaller commission, but it is better to be stable, and it is the best choice for accumulating assets in the early stage.

Now that the funds are not worried, some big cakes should be eaten by themselves.

"Brother! You've finally figured it out! With this technology, you must have pulled it out and done it alone, and there are not a few people like brothers who are willing to give you their lives with a little equity! Wang Kai was overjoyed, and the opportunity for wealth and freedom fell from the sky.

"However, if you start a business, you must be fast, and you must win the project within two or three months." Meng Fanqi has had too many technologies and articles to publish in the past two years, and it is too useless to focus too much on the company's projects, and he also lacks enough senior management experience.

First win a few large projects with absolute technical advantages, and then entrust professional company managers to be responsible for management.

"I can invest in setting up a company, recruit some non-technical people, and make the product to test the waters. You can pay more attention to this matter, and if the subsequent key goals can be achieved, the initial members have a certain percentage of the initial shares. ”

"What was our first major technology product?" Wang Kai was overjoyed at first, but he quickly calmed down.

Although the work of whiteness is a little tiring now, it has been on the rise. It must be right to defect to Meng Fanqi, but you have to ask about the situation clearly, and you can't be too reckless.

"The first core technology is a face recognition algorithm that surpasses human standards." When Meng Fanqi said this, there was no need for superfluous explanations, Wang Kai had already understood the meaning of Meng Fanqi's words.

After all, during this time, he knew a lot about the content of government projects, and knew that high-precision facial recognition systems were a very important government requirement.

"The face detection algorithm on the market is still the old way, and I believe that with your level, the product you make will definitely be able to shake off a lot of those things." Wang Kai suppressed the excitement in his heart and let himself think as calmly as possible.

"The demand for face recognition algorithms is no less than that of road traffic, and it has a wider range of applications in many more fine-grained scenarios. Whether it is a government project such as customs and safety supervision, or a specific punch card to the company, or even mobile payment combined with this year's fire. ”

The more Wang Kai spoke, the more excited he became, and Meng Fanqi's eyes darkened when he heard the company's punch in, and he recalled the unbearable years.

The application of the technology developed by himself to the matter of face punching must be deducted.

"But this matter is not so urgent, it still needs a few weeks of preparation, you first do the work at hand in Baidu, and this part of the Skynet project. If you really want to come out and start a business with me, do a good job of handover as soon as possible, and don't affect the whiteness. After all, Wang Kai is an employee of Li Yanhong, and in all fairness, Li Yanhong has taken good care of himself during this time.

Therefore, Meng Fanqi said a few more words: "If you want to find someone from your other relationships, classmates and friends who used to learn AI, you have the idea to start a business with me, it's best not to pull away the people from the Baidu AI Research Institute." ”

Although the earliest batch of people from the Baidu AI Research Institute basically ran out to do autonomous driving two years later, Meng Fanqi is still reluctant to take the initiative to dig up these AI talents.

On the one hand, it is not good-looking, and on the other hand, he plans to get a lot of data from Baidu in the future.

"I can give this technology out in about a year, and you don't have to rush to give me a reply if you are involved."

"Why don't you hurry, I'm in a hurry, I'll definitely follow you!"