Hacker: A Perfect Start to Surprise

Chapter 489 Create an advanced audio recognition engine! (subscription, custom)

Theoretically speaking, the human vocal cords have been developed and stereotyped after the age of 25.

Although in the days to come, people’s vocal cords may have some subtle changes due to a cold or long-term talking, but in terms of overall frequency and pitch, it can be considered that they have stabilized!

For this, Wang Xiao has a clear understanding.

He learned this theory from biology textbooks when he was in school.

This also means that, in fact, a person’s voice can also be used as a unique basic attribute of a person!

The so-called ignore the person, listen to the voice first.

By listening to the sound, you can tell who this person is!

Therefore, for the audio tracing task, although no good results have been obtained in the dark web.

But now, once I have the massive resources of 20 million cameras, I can play a huge role immediately!

Just do it.

Wang Xiao did not hesitate.

He came to the laptop and clicked on the backstage of the Skynet Project.

Then I logged in to the dark web again, and got the 960 audio files extracted from the audio tracing.

At the beginning, this audio file was extracted from the three videos in that U disk!

It also uses some advanced algorithms to calibrate and repair this audio.

So far, inside this audio, the mysterious man’s voice is already very clear!

After getting this audio file.

Next, Wang Xiao will use a series of very high-end technologies to process this piece of audio information.

He turned on an audio processing software that came with the laptop.

Successfully extracted this audio, the relevant voice features of the mysterious man when speaking, even the tone and tone of voice!

After extracting these key characteristics.

Wang Xiao immediately began to build a mathematical model to mathematically describe the attributes of these tones!

This modeling process, to put it bluntly, is to convert these audio properties into mathematical language.

This is the so-called sound parameter!

This modeling process is also very technical.

Fortunately, when Wang Xiao was in college, he built a similar model in the speech laboratory.

And also wrote a paper!

This also means that from this piece of audio, the pitch, tone, tone and other related parameters of this mysterious man can be extracted!

And convert it into mathematical parameters!

This process sounds very tall.

But under Wang Xiao’s operation, almost less than 20 minutes, all the tones, timbres, and tone samples have been completed!

With these sound parameters, it is not over yet.

Wang Xiao must also find out which cameras have audio recording functions from the 20 million cameras he controls.

For him… it’s not that difficult!

Because these cameras have been completely controlled by themselves.

If you want to know whether it can store sound, it is not easy to do?

He quickly wrote a script recognition program (bcae) sequence to batch recognition.

With the powerful computing power of this laptop.

Half an hour later, from these 20 million camera systems, a camera device with audio storage function was finally screened out!

Wang Xiao looked at it roughly, and it turned out that there are about 6 million cameras with audio storage capabilities!

This means that through the monitoring of these cameras, you can not only see the real-time picture, but also hear the sound coming from the monitoring picture!

With this screening, the next step is quite easy.

He immediately wrote another voice recognition engine program.

This program is difficult to write. It involves statistics, probability, sound physics, sound source field theory, and a series of very high-end knowledge.

But these are nothing to Wang Xiao!

After this voice recognition engine is written.

Wang Xiao then connected the audio parameters of the man just extracted with these more than 6 million cameras!

From now on.

As long as the sound from these cameras matches the sound I just refined.

Then the system will immediately give itself a reminder, and successfully lock the person who made this sound!

As long as the matching success rate reaches more than 75%, it can be determined that the person making the sound may be the mysterious man in the bar!

This technology is very high-end, but the current applications are not very wide.

But in the eyes of a super hacker, he must make some advanced technology in order to achieve his goal.

And use these very advanced technologies to complete the revenge of my sister!

I saw that the voice recognition engine I wrote was successfully deployed and running well.

In Wang Xiao’s eyes, there was an obvious excitement!