This technology can be called a “beauty photo mirror”, which makes the “photo trick” appear its original shape instantly

Sex reassignment surgery in Thailand, cosmetic surgery in Korea, cosmetic surgery in Japan, and facial beauty in China are also known as the four major "witchcraft" in Asia.

Your goddess or male god may be like this…

"Spoofing" is really harmful.

Disguised photos are not small

"Photo tricks" reveal the stuff, and most people laugh it off, but editing or disguising photos can really deceive many people.

There is a "non-existent person" website abroad. Every time it is refreshed, a non-existent face photo is automatically generated, which is extremely real.

These photos are all generated by algorithms, and the naked eye can hardly find flaws. Some are modified with real photos, and some are patched with different facial features of real people.

If you put any one of these photos on the Internet and compile some details of his life, most people may believe that this person really exists.

Compared with the "photo deception" that can be dismantled by looking closely, this almost half-true disguise is difficult to see through and is more harmful.

In 2017, a foreign user used Deepfake technology to graft the face of "Wonder Woman" starring Gal Gadot to an adult movie heroine, and uploaded the video to an adult website, which was strongly criticized by the society Many people worry that they are the next victim of this technology.

It's getting harder and harder to defraud

As technology advances, fake photos and videos are becoming more and more difficult to see through.

For example, the Deepfake technology mentioned above , the algorithm is becoming more and more mature, the interface between the face and the head is more natural, and different expressions can be simulated according to the content of the speech. It has become one of the preferred technologies for fake news. It is difficult to detect with the naked eye .

Before, some mobile phone manufacturers were exposed to SLR photos posing as mobile phone photos for publicity. The reason for the "rollover" is that the netizen checked the EXIF ​​information of the photo, which clearly marked the model of the SLR camera. Soon after, when some mobile phone manufacturers were recruiting employees, it was one of the necessary skills to delete or modify the EXIF ​​information of the photos.

At present, more and more smartphones are taking the route of computational photography, which automatically "optimize" photos, and some manufacturers even directly modify key information in the photos, such as adding texture to the moon, which has caused a lot of controversy.

Recently, some netizens found after using PS to detect that in the photos shown on the official Weibo of a certain mobile phone brand, the aircraft was from the PS. From the time the photo was sent to the time it was exposed, quite a few netizens failed to identify that the photo was tampered with. Yes, they also like it.

"Seal" photos and videos

Seeing is believing has been overthrown again and again in the Internet age, and the human eye can easily be deceived by disguised or processed things.

Is there any way to completely solve the problem of "photo fraud"?

A startup company in San Diego, Truepic, developed the Truepic Foresight function.

When taking a photo, the user can switch the camera to the Secure mode, and a security "seal" will be added to the captured video and audio files , "If someone wants to modify any part of the photo, such as pixels, date, time, or location …Basically, the seal will be destroyed, and we can know that this file has been modified since the filming,” said Sharif Hanna, Truepic's vice president of research and development.

In order to improve the security of this product, the Secure mode will call the 3D depth-of-field camera to record the depth information parameters. "Therefore, you (the user) have complete and transparent information about what happens after the light enters the camera." As long as someone tampered with the photos and videos, It is easy to be found.

Old photos can also be identified

Is it impossible to be sure that a photo or video taken without the Secure mode has been tampered with?

At the same time, Truepic has developed a picture identification function that can monitor whether pictures have been tampered with from multiple dimensions such as pixels, positioning, and time.

In order to solve the pain points in the business field, Truepic also developed a Vision app for point-to-point secure photo transmission and verification.

For example, if company A needs company B to take photos of some documents and products to send to itself, it can send a designated Vision app download link to company B. The list of files and photos is already listed in the app. After company B installs the app, follow the list requirements Use the app to take a photo, and the photo will be automatically uploaded to the cloud. Company A can perform a security check from its Vision app to see if the seal is broken. In this way, multiple security guarantees from shooting, transmission to viewing are realized.

How to use "seal"?

Truepic said that Truepic Foresight technology is initially integrated into a mobile phone equipped with a Qualcomm Snapdragon 865 processor. When the camera is switched to Secure mode, the captured photos or videos will immediately move from the image sensor to a trusted isolation environment. , Generate a digital signature and put a security "seal" on it.

Truepic said it is cooperating with Microsoft, Adobe and media organizations to develop open source software, hoping to popularize it in various hardware and software.

If this technology is popularized, I believe it will be a very big blow to "photographing people", and it will also cause some troubles. Is the post-photographing considered tampering? Is the change of selfie facial appearance caused by the correction of photo distortion a fake?

Is the "seal" really safe?

Sharif Hannah, the vice president of research and development at Truepic, once said such a sentence during the introduction:

After hearing this sentence, I felt strange, and vaguely felt that something was wrong.

The history of the development of the Internet proves that no technology or code is perfect. If the "seal" technology is cracked, the "seal" can also be faked. Will it cause greater harm?

It's like an obscene video that uses DeepFake technology to put a person's face, and this video is "seal". After the overwhelming public opinion criticizes it, what is the use of returning the person's innocence?

#Welcome to follow Aifaner's official WeChat account: Aifaner (WeChat ID: ifanr), more exciting content will be provided to you as soon as possible.

Ai Faner | Original link · View comments · Sina Weibo