link: https://youtu.be/S951cdansBI?si=sU3a2FofTlPiTNqv
Лучшая стратегия освоения материала:
- послушать мое объяснение на ютубе;
- перечитать текст, повторить новые слова;
- многократно переслушать оригинальный отрывок.
Видеоурок
Transcript
4:38
We now have the ability to make people look like other people. Like, I could be Johnny, or I could be Nick, the studio manager. Look, I’m Nick the studio manager right now. I could also be Tom, the music composer. And what’s crazy is even if you go back and watch what I just showed you frame by frame, you likely won’t be able to tell.
Let me explain how on earth we did this. We can do all of this because of this clever computing process. It’s called Generative Adversarial Networks, or GANS. This is where two AIs work in tandem, to get the best fake image possible.
So you have these two AIs. One of them is a forger, and the other is a detective. The forger creates an image based on what you ask it to make, and then it shows that image to the other AI, the detective. The detective goes over the image, and points out all of the reasons it’s fake. It knows what to look for, because it’s been trained on hundreds, sometimes thousands of images of exactly what the finished products are supposed to look like.
The forger is like, okay, cool. Let me try again. And it goes away and it makes another image, fixing the pixels that the detective AI pointed out. Then it shows it to the detective. The detective AI once again points out all of the weaknesses in the fake image, and the forger goes back and makes a new one. And over and over and over again.
And so the longer you give one of these AI models to train on, it basically goes through this creation and improvement cycle, over and over, until you get the best deepfake possible. There are many ways to train a deepfake, but this is the most common, and it’s actually how we got the deepfake that we’re running on this clip right now.
My face is slowly morphing into something else, and it’s basically pixel perfect. We had to run a training model on this face for two weeks. But it got really good. Look. It’s like, amazing. I’m not me. I mean I am me. But I’m not me to you. And that’s kind of nuts.
06:30
Примеры к некоторым словам
Frame: A single, continuous still image in a sequence making up a video or film.
- Example 1: I examined the video frame by frame to find the error.
- Я изучил видео кадр за кадром, чтобы найти ошибку.
- Example 2: The frame rate of the game was excellent.
- Частота кадров в игре была отличной.
To tell the difference: To distinguish between two or more things by comparing them.
- Example 1: It’s hard to tell the difference between the twins.
- Трудно различить близнецов.
- Example 2: She is so skilled that she can tell the difference between real and fake diamonds.
- Она так искусна, что может различить настоящие и поддельные алмазы.
How on earth: A phrase used to express extreme surprise, confusion, or disbelief about something.
- Example 1: How on earth did you solve that complicated problem?
- Как черт возьми ты решил эту сложную задачу?
- Example 2: How on earth did you get here so fast?
- Как черт возьми ты здесь оказался так быстро?
Adversarial: Pertaining to conflict, opposition, or rivalry.
- Example 1: The two companies have an adversarial relationship.
- У двух компаний конфликтные отношения.
- Example 2: He is known for his adversarial approach to politics.
- Он известен своим конфликтным подходом к политике.
In tandem: Working together closely and efficiently.
- Example 1: The two departments worked in tandem to complete the project.
- Два отдела работали совместно, чтобы завершить проект.
- Example 2: The engine and the transmission operate in tandem for optimal performance.
- Двигатель и трансмиссия работают совместно для оптимальной производительности.