This AI app enhances your crappy photos — or turns them into monstrosities

AI is having a big impact on photo editing, but the results are proving divisive .

The proponents say that AI unleashes new artistic ideas and cuts the time creators spent on monotonous work. Critics, however, argue that the techniques distort reality and propagate an artificial homogeneity.

A new entry to the debate is GFP-GAN. The system gives low-quality images a high-resolution revamp — and the results can be impressive.

GFP-GAN’s outputs (far-right) look impressive compared to other state-of-the-art techniques. The system, which was created by researchers at the Tencent ARC Lab in China, uses a generative adversarial network (GAN) architecture to enhance faces in old, damaged, and unclear photos. The features are then restored and refined to generate an upscaled image.

“While previous methods struggle to restore faithful facial details or retain face identity, our proposed GFP-GAN achieves a good balance of realness and fidelity with much fewer artifacts,” the study authors wrote in their paper . “In addition, the powerful generative facial prior allows us to perform restoration and color enhancement jointly.”

The research team has also kindly created a demo for their system — which gave us a chance to put their claims to the test.

In our brief and highly unscientific trial, GFP-GAN was most adept with images that had only minor blemishes. Unlike many AI photo tools , it also generated a similar accuracy across a diverse range of subjects. The problem was, this accuracy was highly inconsistent.

Unsurprisingly, this issue was most pronounced when the tool was used on extremely blurry images. However, even high-resolution inputs led GFP-GAN down some horrifying uncanny valleys.

Our experiment left me on the fence in the debate over AI photo editing. I can certainly see the creative potential, but I certainly wouldn’t trust it to reflect reality.

Perhaps that isn’t such a bad thing. As privacy professionals have warned , cameras equipped with AI image enhancers could be used to track entire populations.

Apple wins tax battle against EU, avoids $15B fine

Apple has won its appeal against an EU order to pay Ireland $15 billion in back taxes, after Europe’s second-highest court rejected the bloc’s claims that the tax breaks were an illegal subsidy.

The case stems from a 2016 European Commission ruling that Ireland had breached the EU‘s state-aid rules by giving Apple illegal tax benefits for more than a decade.

The EU‘s General Court on Wednesday annulled the decision because the Commission hadn’t shown “to the requisite legal standard” that Apple had received an unfair advantage.

“The General Court considers that the commission did not prove, in its alternative line of reasoning, that the contested tax rulings were the result of discretion exercised by the Irish tax authorities,” the court said in a press release .

Apple responded to the decision in a statement:

“This case was not about how much tax we pay, but where we are required to pay it. We’re proud to be the largest taxpayer in the world as we know the important role tax payments play in society.”

The Irish government had joined the firm in appealing the ruling, as its low tax regime has attracted 250,000 multinational companies.

“Ireland has always been clear that there was no special treatment provided to the two Apple companies. The correct amount… was charged in line with normal Irish taxation rules,” the finance ministry said in a statement.

However, opposition parties slammed the ruling. Sinn Fein finance spokesperson Pearse Doherty said: “Today was a bad day for the Irish taxpayer.”

The decision is also a big blow to Margrethe Vestager, the European Commissioner for Competition. Vestager had led the probe into Apple’s “sweetheart deal” and has made crackdowns on low-tax regimes the focal point of her time in office.

How long until ‘digital humans’ take over dating apps?

In the early months of the COVID-19 shelter-in-place, Cathy Gover joined the dating website Plenty of Fish . Like many people during the pandemic , the Tennessee widow was suffering from the onset of lockdown loneliness and decided to look for love online. It wasn’t long before she met Marc from Atlanta, and fell for his considerable charms. Then, six weeks into their online romance, Marc asked Cathy for financial help .

Of course, you’ve heard this story many, many times before. An older woman looking for love and companionship meets a predator posing as a lonely heart, only to be duped out of thousands of dollars. Sometimes these cases can be frustrating, and leave us asking how the victim missed all the glaring red flags.

In the end, Cathy discovered that the pictures Marc had sent her were actually of a Brazilian pastor. But not before he had relieved her of a cool $3,000 and badly damaged her faith in humankind and, of course, her dampened hopes of finding love again.

With just a few stolen photographs and a six-week “ love bomb ” operation, this fraudster was able to extract a significant amount of cash from Cathy and he (if it even was a he) has no doubt done the same thing to scores of others. Now, just imagine how many more he could fool if he was able to create hundreds of original, convincing, and interactive identities at the click of a button.

This scary prospect isn’t as far off as you might think.

Enter the MetaHumans . High-fidelity digital humans that can be created “in minutes” with the help of a new tool from Epic Games — the MetaHuman Creator — that promises to “ will empower anyone to create a bespoke photorealistic digital human, fully rigged and complete with hair and clothing .”

Though not yet perfect, these digital creations are still pretty uncanny and demonstrate just how far this kind of technology has come.

Obviously, Epic’s primary audience for this impressive tool is creators in the gaming, movie, and wider entertainment industries, and we can only imagine what fantastical experiences they will use to create, but we also shouldn’t rule out its malicious use. Particularly in this new, more virtual world that often relies on remote video communication more heavily than in-person interaction.

As a species that has thus far evolved to trust our senses, but we need to slowly unlearn the idea that “seeing is believing.” At the same time, more serious thought should go into how we get ahead of a future in which legions of these MetaHumans (and their equivalents) are unleashed online to mislead and defraud those who cannot identify them as fakes.

But even if governments were able to regulate in a way that would block or deter the weaponization of this technology (which seems unlikely), we should also take pause to think about how we feel about the legitimate, transparent use of these phony humans outside of the narrow world of movies and video games.

Might they replace humans in commercials, or even in the classroom? Would you have one read your kids a bedtime story? Could they model clothes without tripping up on virtual runways, or host the next online conference you attend without a hitch? There are many possibilities, but each heralds the redundancy of a real human being. It makes you wonder why we are so hell-bent on creating tech in our own image when it could be the ultimate act of self-sabotage…

Even if we choose to dismiss as fantasy the idea that digital humans will infiltrate our daily lives in the near-term, we should not overlook that for much younger generations concepts like this — along with fake news and deepfakes and tech like GPT-3 — will be part of their understanding of the world. They are altogether less “creeped out” by AI, and its creations, and more likely to engage with bots in all guises . As such, we have a real responsibility to think about how we want these tools to be deployed and which lines we should prevent them from crossing… even if it does all feel a little Dr. Who .

This article was originally published on You The Data by Fiona J McEvoy . She’s a tech ethics researcher and the founder of YouTheDataom .

Leave A Comment