Discuz! Board

 找回密碼
 立即註冊

QQ登入

只需一步,快速開始

搜尋
熱搜: 活動 交友 discuz
檢視: 197|回覆: 0

What are deepfakes?

[複製連結]

2

主題

2

帖子

8

積分

新手上路

Rank: 1

積分
8
發表於 2023-10-5 13:31:52 | 顯示全部樓層 |閱讀模式
Understand what deepfakes are and what dangers these actions pose. article banner Share: Do you need innovation and digital transformation for your business? Find the best experts here at SIS Innov & Tech! Talk to Our Team The internet is used by most people, whether in a positive or negative way, and for a multitude of things. Every day, new discoveries, new terms, new practices and new ideas emerge in this space where, often, the sky is the limit. One of these creations was the deepfake. What is it? Deepfake is a technology that uses artificial intelligence to create fake but realistic videos that put people in situations that never happened. Examples of deepfake content are adult content featuring celebrities or videos of influential politicians giving fictitious speeches.
Deepfakes are capable of reproducing the mannerisms, appearance and even voice of someone in the real world. The name is derived from the combination of two expressions in English: deep learning and fake. Deep learning is an evolution of artificial intelligence improvement methodologies. It is a derivation of machine learning . The concept came along with the first advances in artificial intelligence, around 1950, and literally means putting the computer to learn. The idea is to make the electronic brain study algorithms to understand how to read data and make decisions more accurately.
It works as if he watched a film several times and could reproduce what he learned, however, creating a completely new story. You might like: Do you know what the Millennium Bug was? image of the author By: SIS Team October 5 , 2023 read more article cover How are deepfakes created? Deepfakes are created in two steps: in the first, the software whatsapp number list   captures reference images of the person who will be used in the video. This way, the more images, the more precision we will have. For more reality, it is interesting to create an image bank with different angles and expressions. The second part is time to record the movements of a second person, which will serve as a base. Now it's artificial intelligence's turn to play its role and unite the two to form the deepfake. Check out some examples here.

You might like: How to protect children on the internet? image of the author By: SIS Team October 5 , 2023 read more article cover The history of deepfake The term deepfake emerged in December 2017, when a user on the website Reddit, using the name “deepfake”, posted fake sexual videos with celebrities. Using deep learning software, he applied familiar faces to videos that already existed. The most notorious cases were those of actresses Gal Gadot and Emma Watson. Thus, the expression deepfake began to be used to indicate videos edited through machine learning and other artificial intelligence capabilities. In January 2020, Facebook released an announcement on its blog stating that it would ban deepfake videos from its platform, as this type of content manipulates reality and is a major challenge for the technology industry. Videos are banned when the edits are not so obvious to users or when the content leads the public to believe that a certain thing was said when, in reality, this did not happen.
To carry out these checks, Facebook has 55 checkers, who speak 45 different languages ​​to analyze and flag fake videos. In September 2019, Facebook donated 10 million dollars to a fund that aimed to improve technologies for detecting materials like these. However, at one point, the company received criticism for not removing a fake video of the president of the US House of Representatives, Nancy Pelosi, which went viral in the country in 2021. Deepfakes can just be a prank video without causing significant harm. But, in many cases, it can be an instrument for the dissipation of fake news, causing misinformation and manipulation. In fact, Mark Zuckerberg himself was one of the victims of deepfake software.

回覆

使用道具 舉報

您需要登入後才可以回帖 登入 | 立即註冊

本版積分規則

Archiver|手機版|小黑屋|DiscuzX

GMT+8, 2024-11-27 04:04 , Processed in 0.020084 second(s), 18 queries .

Powered by Discuz! X3.4

Copyright © 2001-2020, Tencent Cloud.

快速回復 返回頂部 返回列表