Computerphile’s Rob Miles takes a closer look at GPT-2, the AI deemed “too dangerous” by its creators to release. If you’ve not heard about it, it’s an AI that, given a bit of text to prime it, it continues writing a believable and coherent way.
Two Minute Papers shows off the fascinating work in the paper “Few-Shot Adversarial Learning of Realistic Neural Talking Head Models.” Could this be the next evolution of GANs? It certainly will empower a whole new wave of deep fakes.
What a time to be alive, indeed!
Generative models, commonly referred to as GANs, are a family of AI architectures whose aim is to create data samples from scratch. They work by capturing the data distributions of the type of things we want to generate.
Here’s an interesting read on the topic.
These kind of models are being heavily researched, and there is a huge amount of hype around them. Just look at the chart that shows the numbers of papers published in the field over the past few years:
Here’s an interesting video and project (source code) to generate human faces with AI.
In this TED talk, you will meet AIVA, an artificial intelligence that has been trained in the art of music composition by reading more than 30,000 of history’s greatest scores. In a mesmerizing talk and demo, Pierre Barreau plays compositions created by AIVA and shares his dream: to create original live soundtracks based on our moods and personalities.
Two Minute Papers explores the paper “Vid2Game: Controllable Characters Extracted from Real-World Videos” and how you can put yourself into the game.
Two Minute Papers explores the paper “Direct speech-to-speech translation with a sequence-to-sequence model.” Check out the voice samples as well.