1995, Starting as a Criminal Investigation Expert

Upon arriving at Peking University, Lao Zhou showed his ID, drove into the campus, and headed straight for the School of Mathematical Sciences.

Another advantage of riding in their car is that they both have identification documents, making it very convenient to go to many workplaces.

Wei Ranhui saw everything, and he knew that the drivers in the station didn't have such solid credentials.

But Li Shu didn't say anything, and he didn't ask any more questions.

The car was parked on the side of the road, and the two got out and walked towards Jingyuan No. 1 Courtyard.

Wei Ranhui looked around and sighed, "It's been more than ten years since I graduated in '82, and this seems to be the first time I've returned to Jingyuan."

Li Shu asked curiously, "Director Wei, I remember you were from Tsinghua University, right?"

Although Tsinghua University and Peking University are next to each other.

Wei Ranhui nodded and said, "That's right, I'm from the Computer Science Department at Tsinghua University. But when I was in school, the Computer Science Department at Peking University had just been established, and we came here quite often."

He pointed north and said, "Back then, they were in the South Pavilion, which was just across the road from the First Courtyard, right next to it."

As she spoke, a smile appeared on her lips, as if she were recalling something.

Judging from his reaction, Li Shu guessed that he was probably thinking of some junior or senior students.

Otherwise, why would someone from Tsinghua University keep going to Peking University for no reason?

When I arrived at Geng Zheng's office, he was there, talking to a tall man wearing glasses, probably a graduate student.

Li Shu overheard the conversation as he entered and felt that the two were talking about something completely foreign to them.

Upon seeing Li Shu, Geng Zheng's eyes lit up, and he quickly invited the two of them to sit down and asked the student to make tea.

After Li Shu put down the gifts, he hurriedly pushed them out, saying, "We're so familiar with each other, just tell me if you need anything. I'll definitely help if I can."

Li Shu felt he couldn't explain himself clearly, so he asked for paper and pen and wrote down a model formula.

Geng Zheng took it, glanced at it, and felt a vague recollection of it. He frowned for a moment before relaxing, then casually handed it to the tall student next to him, asking, "Xiao Guo, what's this?"

The tall student took it, glanced at it, and his expression immediately froze. He quickly looked at Li Shu and said awkwardly, "Teacher, this looks like points?"

Geng Zheng shook his head slightly and asked in return, "Li Shu, this is the F model, right? What's wrong? Why did you bring this up?"

Li Shu nodded to himself, thinking to himself that a professor, after all, had a wide range of knowledge.

The model I wrote down is called the f-model. It was a mathematical model proposed by three American experts in 92.

The reason it's called the F model is because it's an abbreviation of the names of the three people involved.

Li Shu explained, "Professor Geng, Director Wei and I are preparing to study image restoration technology, and we plan to use this model as a basis to complete practical applications."

:

Li Shu then explained his ideas.

The mathematics part will be handled by Geng Zheng and his team, while the software implementation will be completed by Wei Ranhui and his team.

Geng Zheng glanced at Li Shu, his eyes looking somewhat strange.

He had a lot of contact with Li Shu, and of course he knew that Li Shu had made significant innovations in many areas of criminal technology and was an incredible genius.

But he knew Li Shu's mathematical skills all too well; they couldn't be described as bad.

Although the f-model is in the field of image processing, its core is definitely still mathematics, belonging to the field of applied mathematics.

Although the model proposed by the three is quite cutting-edge, it is not that difficult in terms of mathematics itself.

But even if it's not high, it's still beyond Li Shu's reach.

After thinking for a moment with his arms crossed, Geng Zheng said, "Li Shu, I understand your thoughts. But how should I explain this to you?"

"Let me give you an example. It's like a disease. The f model tells us how to treat the disease, which seems to solve the problem, but in reality, it doesn't."

"We have the method, but we need to make a medicine. We don't know how to make this medicine, or even if there is a way to make it, it's too difficult and impossible."

Do you understand what I mean?

Li Shu nodded, understanding perfectly.

Based on the systematic data, the f model, which was proposed in 92, took more than a decade of research to gradually mature.

Li Shu knew exactly what the key to maturity was.

Before Li Shu could speak, Wei Ranhui interjected, politely asking what the specific problems were.

Geng Zheng said, "It seems to be a problem with the calculations, but I don't really understand the specifics."

I felt a little embarrassed after saying that.

He was just using metaphors and all that, but then he said he didn't understand.

But in reality, they just don't understand.

After all, everyone has their own area of ​​expertise. I specialize in causal probability, while the f model requires experts in partial differential equations and finite difference equations.

However, whether this is the School of Mathematical Sciences or Peking University, I may not understand it myself, but many others do.

He hesitated for a moment, then beckoned Li Shu to go outside to call for help. As they walked, he asked, "Li Shu, although I don't know much about this, I know it's not easy. Just coming to us isn't enough; we might not be able to help you solve it."

When working on the probabilistic model last time, Li Shu first presented a completely new model with a framework and methodology, and then completed the specific algorithm himself.

In other words, Li Shu provided the solution, while he himself was just a calculator.

:

But this time it's different. The f model has been proposed for four years, and not only internationally, but even the proposer himself hasn't figured it out and doesn't know how to implement it.

As a result, Li Shu wanted the mathematics department to help him sort it out, but that's not something that can be done just by talking.

Li Shudao said, "Professor Geng, I understand what you're saying, but don't worry, since I dared to come, I must have thought it through. I'm here to break the ice."

Geng Zheng was taken aback. Should he poke a hole in the window?

He led Li Shu up to the second floor, to the Department of Scientific and Engineering Computing.

After the merger of the College of Mathematics, there are four departments. The predecessor of the Department of Mathematics and Engineering Computing, the Computational Mathematics Teaching and Research Section, was established in 55 and has always been a key department.

Jingyuan is small, and many of its buildings are large offices.

Geng Zheng peeked over and asked Li Yuanping where he had gone.

I learned that it was right next door, in a small room, probably a small office, filled with all sorts of books.

Two people were standing in front of the activity blackboard, which was covered with characters that Li Shu couldn't understand. They must have been discussing something.

"Professor Li, I'm busy." Geng Zheng knocked on the door.

One of the men, with a long, thin face and a refined demeanor, turned his head; he must be Li Yuanping, whom Geng Zheng was looking for.

I greeted them.

Li Yuanping is now an associate professor in the Department of Computer Science, and he is 42 years old this year.

The other, slightly younger man, named Hu Anming, is only 32 years old and is currently a lecturer in the Department of Applied Mathematics at Tsinghua University next door.

With Wei Ranhui added to the mix, Li Shu realized that everyone in the room except himself was from Tsinghua or Peking University.

Both of them were very polite.

Wei Ranhui was a graduate of the first class in 82 and was also the head of the department, so he had sufficient qualifications.

That's Li Shu, now an associate researcher and deputy director.

Geng Zhengdao: "Professor Li, I remember you seem to have studied the f-model?"

Li Yuanping's main research focus is now on numerical algorithms for variational problems, and the f model involves this area.

Li Yuanping poured tea for everyone and nodded, saying, "That's right, I have done some research. What's wrong?"

Geng Zhengcai pointed to Li Shu and explained Li Shu's thoughts.

Li Yuanping glanced at Li Shu and the other person with surprise.

:

Wei Ranhui is from the Computer Applications Department and works on video restoration, which is his field. As for f, which is cutting-edge research in image processing, it is reasonable for him to want to explore it and apply it in practice.

Li Yuanping pondered for a moment and said, "Director Wei, your idea is good, no problem, but this matter is rather difficult."

Hu Anming, who was standing next to him, nodded as well, clearly aware of the situation.

Li Yuanping simply got up, flipped over the blackboard, first wrote down the formula for model f, and then pointed to it, saying, "The model has been proven to be very perfect. The key is how to solve this on a computer."

"Initially, the paper used gradient descent, but the computational cost was too high. In recent years, some people have been researching and trying to find a smoother processing method, as well as iterative algorithms. By the way, Anming, I think I saw a paper this year that said they were going to start with fixed-point iteration."

Hu Anming nodded slightly and chimed in, "Simply put, according to the original algorithm, it might take several days to process just one image."

"Now that we're using fixed-point iteration, the time has been reduced by several times, maybe ten hours or so, but it's still very slow."

A single image, even if it's only one or two hours old, is still useless, let alone ten or more.

Li Yuanping nodded in agreement: "That's right, that's the point. This model is very good, and the academic community believes it has significant research value, but its shortcomings are also obvious. At least for now, the problem of implementation is far from being solved."

The subtext is that it's no use you contacting us.

Cutting-edge theoretical research and major computational challenges are not things that can be solved overnight.

Wei Ranhui turned to look at Li Shu and gave him a look.

Since you brought it up, isn't it your turn to perform?

Geng Zheng also turned his head and looked over, giving him the same look that said, "You're not going to say anything."

What you just said, is about breaking through the window paper.

Li Shu chose to ignore it and said with a smile, "Professor Li, Teacher Hu, after hearing what you said, I realized that there is such a big problem here."

First of all, Li Shu had absolutely no concept of the issues the two were discussing, such as gradient descent and fixed-point iteration.

Li Shu could only understand the general descriptions of the information in the system.

For example, a problem was encountered at the beginning, and then someone suggested a solution.

Li Shu still couldn't understand how it was actually resolved.

Unfortunately, this is essentially a mathematical problem.

And mathematics...

Li Shu now has an extraordinary memory, able to memorize anything quickly, and his learning ability has improved significantly, but he still struggles with mathematics.

"But you two may not know that methods like Wiener filtering, which we commonly use now, are too ineffective, and we do need a new image processing technology."

The two exchanged a glance and nodded slightly.

:

Li Shu continued, "Well then, could you two find us some relevant research materials or journal directories? At least we can get a deeper understanding and see if it's feasible."

I said I was here to "poke a nutshell," but it was just a casual remark.

When Li Shu came, he had already thought it through. This was not like fingerprint extraction, or even NA typing, or the field of chemistry. He had a flash of inspiration, and his thoughts flowed like a spring, resulting in a brilliant idea.

This is mathematics, which is the most logical thing.

If I were to directly write out the derivation formulas and solution steps for the dual projection algorithm from the system documentation, it would definitely scare Li Yuanping and the other man to death.

Even Mendeleev could never have dreamed of such a thing.

Therefore, Li Shu considered laying the groundwork by first obtaining some information and conducting research before proposing a solution that would at least be more reasonable.

Based on the available information, this dual projection algorithm is also an extremely ingenious solution, much like developing your own magnetic bead coating or magnetic solid-phase extraction.

Mathematical tools have existed for a long time, but no one thought of connecting them to solve the calculation problem of model f.

Li Yuanping and Li Shu exchanged a glance, their eyes filled with doubt. But since Li Shu had brought it up, and it wasn't a troublesome matter, they nodded and said, "If you want to know, of course, that's no problem. If you have any questions in the future, you can come to me."

Li Shu seized the opportunity and said with delight, "That's wonderful, Professor Li. Could you please leave me your contact information so we can easily reach you if we have any questions later?"

Li Yuanping thought to himself, "You two are really going to do research, huh?"

One is a computer science major, and the other is someone whose major is unclear, both researching these kinds of mathematical problems?

But since he had already asked, he couldn't refuse. So he left his contact information and called over a student to compile a list of information for Li Shu.

Wei Ranhui winked at Li Shu, who ignored him, thanked him profusely, and then suggested having lunch together, offering to pay for it as a token of gratitude.

They arrived right at mealtime.

After finishing the meal, I took the three of them back to school, got the roster, and then said goodbye and left.

Watching Li Shu and his companion leave, Li Yuanping couldn't help but ask, "Geng Zheng, what do they mean by this? They're not really expecting us to solve this problem for them, are they?"

If I had that ability, getting a professorship shouldn't be a problem.

To be honest, Geng Zheng couldn't figure out what Li Shu was thinking either. He seemed very confident, but in the end, he didn't say anything.

But after thinking about it, Geng Zheng said, "Professor Li, my suggestion is that you contact him more often. He has an incredible learning ability, and most importantly, he's a genius. He might be able to help you and give you some inspiration."

genius?

Li Yuanping frowned slightly. This is Peking University; who isn't a genius here?

: