Lip-syncing app Wombo shows the messy, meme-laden potential of deepfakes

You’ve probably already seen Wombo videos floating around on social media. Perhaps Ryu Street Fighter Sing “Witch Doctor” or Last Three heads of the U.S. Federal Reserve Imitate Rick Astley’s “Never Going to Give You Up”. Each clip has exaggerated facial expressions and bizarre, sometimes nightmare-like animations. They are stupid and fun, and usefully show the current state of deepfakes.

Creating an AI-generated fake is definitely getting faster and easier, but the more convincing it is, the more work it takes. For example, the realistic Tom Cruise deepfake, buzzed by TikTok, had to be prepared for weeks to become an experienced VFX artist, best aviator. One-click fakes, which you can create without effort and expertise, by comparison, are still similar to those made on the Wombo app, and will continue to do so in the near future. At least in the short term, Deep Fake will definitely be a manipulated and immediate meme bait.

The Wormbo app was released in Canada at the end of last month after a short development process. “I got the idea for Wombo by co-smoking with my roommate in August 2020,” says Ben-Zion Benkhin, app creator and CEO of Wombo. The Verge. It was a “great pleasure” to launch the product. “I’ve found an AI space, a meme space, a deep fake space, and an opportunity to do something cool.” In just a few weeks, Benkhin estimates that the app has recorded about 2 million downloads.

Wombo is free and easy to use. Take a picture of your face or upload an image from your camera roll and press a button to let the image lip sync to one of the songs adjacent to the meme. The app’s software will enchant everything that vaguely resembles even a face, and a lot of things that don’t. Although similar apps in the past have been plagued with privacy fears, Benkhin is determined that the user’s data is safe. “We take privacy really seriously,” he says. “All data will be deleted and we will not share it or send it to anyone else.”

The name of the app is specifically derived from esports slang. Super Smash Brothers close-up. “When the player lands like a crazy combination, the caster is’Wombo Combo! Wombo Combo! ‘” says Benkhin. True to these origins, Wombo has proven to be particularly popular with gamers who have used it to animate characters for titles such as: The League of Legends, Fallout: New Vegas, and Dragon age. “I am [the origins of the slang]Benkhin says: “And there was a pizzeria that started all of this, and I would call it a wombo combo with toppings on every pizza.”

Benkhin says the app works by morphing the face using predefined choreography. He and his team filmed a basic video of each song in his studio (“Really my apartment’s room”) and then used it to animate each image. “We steal motion from their faces and apply them to photos,” he says. The app is also an example of the fast-paced world of AI research where new technologies can become consumer products in weeks. Benkhin points out that the software was built “on top of the existing work”, but with subsequent adjustments and improvements it has been made “our own model”.

Currently, Wombo only offers 14 short song clips that can be lip-synced, but Benkhin says he plans to expand these options soon. When asked if they have an appropriate license for the music they use in the app, he refuses to answer, but says that the team is working on it.

However, like TikTok, the scope provided by Wombo can help ease worries about the rights of the licensee. Benkhin has already approached Wombo by artists who want to get their music out of the app, which has the potential to provide an additional source of revenue for the current premium tier (paying priority processing fees and not paying for in-app ads). “It will give [artists] It’s a whole new way to captivate your audience,” he says. “It gives them a new viral marketing tool.”

Wombo is far from the first app to use machine learning to create fast and fun deepfakes. Others include ReFace and FaceApp. But it’s the latest example that will become an increasingly noticeable trend as Deep Fake apps become the latest meme templates, allowing users to combine their favorite characters, trendy songs, choreographed dances, public figures, and more. The future of deep fakes will certainly be memetic.


Please enter your comment!
Please enter your name here




T-Mobile saw Verizon telling people to turn off 5G and thought ‘why stop there’

T-Mobile is busy tying up 5G networks, and in recent years it has spent billions of dollars extra to expand. A...

Legal chatbot firm DoNotPay adds anti-facial recognition filters to its suite of handy tools

Legal services startup DoNotPay is best known for its "robot attorney", an automated bot that handles tedious online tasks such as...

Samsung’s Odyssey Neo G9 is a high-end TV disguised as a 49-inch curved gaming monitor

The follow-up to Samsung's excellent Odyssey G9 curved gaming monitor is pretty much here. The 49-inch Odyssey Neo G9 will...

Apple WWDC 2021 announced as online-only event

Apple has announced that the Worldwide Developers Conference (WWDC) will be held from June 7th to June 11th this year, and...

Bose Frames Tempo review: the specs to beat

It's still fair to call audio sunglasses a niche category, but it's definitely growing as Bose offers Amazon across multiple models,...