The goal of generative AI is to take training samples from some unknown, complex data distribution (e.g., the distribution of human faces) It’s important to be familiar with deep learning and comfortable with Tensorflow/Keras. The generated data should be realistic and accurate compared to the actual data distribution.
Table of contents
Learn how to generate images using VAEs, DCGANs, and DDPMsIntroduction:Generative Model Trilemma:Variational Autoencoder:Deep Convolutional GANs:Diffusion Models:Conclusion:1 Impression