Use case
Image generation
Produce images from a text prompt, and understand why the two dominant approaches differ.
What it takes
In the order you would adopt them. Every step opens in the knowledge graph.
-
Generates data by learning to reverse a gradual noising process.
-
Generator and discriminator trained against each other to synthesise realistic data.
-
Convolutional neural network, sharing weights across positions to exploit spatial structure.
-
Mechanism that weights each part of an input by its relevance to every other part.