A Generative Adversarial Network (GAN) involves Generator (G) and Discriminator (D) networks whose purpose, respectively, is to map random noise to samples and discriminate real and generated samples. Formally, the GAN objective, in its ori…We compute the IS for both the training and validation sets of ImageNet. At $128 \times 128$ the training data has an IS of 233, and the validation data has an IS of 166. At $256 \times 256$ the training data has an IS of 377, and the valid…
Alec Radford
scientist · 2 mentions across 1 reading
In this course
Radford is a co-author on foundational work in generative models and large-scale vision systems, appearing here in the context of GAN architectures and image synthesis evaluation metrics. The readings invoke his research to establish baselines for how modern generative systems are measured—specifically through Inception Score and resolution-scaled performance on ImageNet. His work anchors the technical genealogy connecting adversarial training to the scaled image generation systems central to contemporary AI aesthetics.
Mentioned in 1 reading
Appears alongside
People mentioned in the same passages — sorted by co-occurrence weight.