AI Paper Atlas

Research map / Image generation / GANs

Incorporating Reinforced Adversarial Learning in Autoregressive Image Generation

Kenan E. Ak, Ning Xu, Zhe Lin, Yilin Wang

ECCV 2020 · Image generation › GANs · approach: Stabilized GAN training · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add reinforced adversarial learning (single reward)VQ-VAE with PixelCNN prior · CelebA 64x64 oracleNLL: 7.55 -> 7.07Improved
add reinforced adversarial learning (single reward)VQ-VAE with PixelCNN prior · CelebA 64x64 oracleFID on oracle data: 7.49 -> 5.17Improved
add RAL (single reward)VQ-VAE with PixelCNN prior · CelebA 64x64FID: 5.35 -> 3.24Improved
add RAL (single reward)VQ-VAE-2 priors · CelebA 128x128FID: 58.07 -> 49.24Improved
add RAL (intermediate reward)VQ-VAE-2 priors · CelebA 256x256FID: 69.37 -> 64.72Improved
add RAL (intermediate reward)VQ-VAE-2 priors · LSUN-bedroom 128x128FID: 47.77 -> 36.47Improved
use intermediate reward instead of single rewardRAL · LSUN-bedroom 128x128FID: 39.86 -> 36.47Improved
use intermediate reward instead of single rewardRAL · CelebA 128x128FID: 49.24 -> 51.91Hurt
use partial generation instead of full generation (ablation)RAL single reward · CelebA 128x128FID: 51.91 -> 49.24Improved
apply RAL jointly on top and bottom priors instead of top only (ablation)RAL · CelebA 128x128FID: 54.82 -> 49.24Improved
apply RAL jointly on top and bottom priors instead of bottom only (ablation)RAL · CelebA 128x128FID: 55.02 -> 49.24Improved
use 4x4 instead of 8x8 discriminator score map (ablation)RAL single reward · CelebA 128x128FID: 53.15 -> 49.24Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy