AI learns to Speedrun QWOP and sets top 10 time
Wesley Liao trained an AI to learn to play QWOP using reinforcement learning and imitation learning and it managed to achieve a top 10 speedrun. QWOP is...
machine learning
Using AI to create baking recipes
The folks at Google used the power of AI to create a cakie (cake/cookie) and a breakie (bread/cookie) recipe. We started off by collecting hundreds of cookie,...
Using AI to turn animated characters into realistic portraits
Motion graphic artist Nathan Shipley has been using a StyleGAN encoder to turn works of art into realistic-looking portraits. Above are the Mona Lisa and Miles Morales...
Creating 3D models of people from 2D images
Continuing humanity's race towards potential deepfake hell, researchers have developed a way of creating 3D models from 2D images using neural networks. The full title of the...
Gollum lip syncs The Scatman using AI Lip Sync
Another day, another video of movie characters lip syncing to pop songs. This time, somebody turned AI Lip Sync on Gollum to make him lip sync The...
Using artificial intelligence to make movies lip-sync Smash Mouth's 'All Star'
Visual effects engineer Jonty Pressinger (who previously fixed The Lion King remake by deepfaking it to look more like the cartoon version) was testing out Wav2Lip and...
Creating the optimal peanut butter and banana sandwich using computer vision and machine learning
Quarantine does weird things to people, and in the case of engineer Ethan Rosenthal it made him develop a system for optimizing a peanut butter and banana...
Creating photorealistic portraits of Roman Emperors using their sculptures
Artist Daniel Voshart has taken it upon himself to create photorealistic portraits of Roman Emperors using machine learning applied to images of their busts. Using the neural-net...
Using internet photos to create 3D reconstructions
Researchers have developed a system for taking photo collections from the internet and turning them into 3D reconstructions. They call it NeRF-W, which stands for Neural Radiance...
Upscaling the oldest recorded video using neural networks
Using an ensemble of neural networks, YouTuber Denis Shiryaev upscaled the oldest recorded video: Louis Le Prince's Roundhay Garden Scene which was shot in October of 1888....
MIT researchers create deepfake of Nixon delivering 'In Event of Moon Disaster' speech
To illustrate the dangers of deepfakes, a team from the MIT Center for Advanced Virtuality created a deepfake of President Nixon delivering the real contingency speech written...
Researchers discover resolution limit to "upsampling" of pixelated faces
In news surprising only to CSI writers, AI researchers have discovered there is an inherent resolution limit to "upsampling" pixelated faces. Duke University researchers have created an...
Algonuts is machine generated Peanuts
Algonuts is the product of techno-artist shardcore, who trained a machine learning network on 18,000 of Charles Shulz Peanuts strips: Charles Shulz, the creator and artist of...
A Website Where An Artificial Intelligence System Will Write Song Lyrics About Your Topic Of Choice
Above: A very sad country song about geeks. These Lyrics Do Not Exist is a website connected to an artificial intelligence system that will write song lyrics...
Jim Carrey Deepfaked As Jack Nicholson In The Shining
This is another deepfake video from Youtuber Ctrl Shift Face, who used machine learning to replace Jack Nicholson's face in The Shining with Jim Carrey's. It's very...
A Glitch In The Matrix: Programmer's First Attempt At Having Cars Removed From Video Footage Using Machine Learning
This is a video of programmer Chris Harris's first attempt at having an artificial intelligence system using neural networks remove cars from some footage he shot (links...