March 2026 Course - Lecture 2: GANs and Early High-Fidelity Image Generation
AI Course
Save $1.00
Original Price:
$7.00
Current Price:
$6.00
Digital products are non-refundable after purchase.
1451 pages of documents. No vids.
Core papers:
Generative Adversarial Nets
Ian Goodfellow et al., 2014DCGAN: Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
Radford, Metz, Chintala, 2015StyleGAN: A Style-Based Generator Architecture for Generative Adversarial Networks
Karras et al., 2018 / 2019
Topics:
Generator vs discriminator
Adversarial training
Why GAN images are sharp
Why GANs were important for face generation and deepfake systems
Limitations: instability, mode collapse, weak controllability
Engineering connection:
Deepfake pipelines
StyleGAN face generation
Super-resolution
Why diffusion models later became dominant