AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

GloTok: Global Perspective Tokenizer for Image Reconstruction and Generation

Xuan Zhao, Zhongyu Zhang, Yuge Huang, Yuxi Mi, Guodong Mu, Shouhong Ding, Jun Wang, Rizen Guo, Shuigeng Zhou

AAAI 2026 · Image generation › Autoregressive and latent generation · approach: Generative latent space tooling · 3 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 histogram relation learning to dual-codebook tokenizer (ablation)GloTok · ImageNet-1k reconstruction ablation (4096+12288 codebooks)rFID: 1.43 -> 0.97Improved
add histogram relation learning to dual-codebook tokenizer (ablation)GloTok · ImageNet-1k reconstruction ablation (4096+12288 codebooks)PSNR: 22.06 -> 22.30Improved
add residual learning module on top of histogram relation (ablation)GloTok · ImageNet-1k reconstruction ablation (4096+12288 codebooks)rFID: 0.97 -> 0.92Improved
add residual learning module on top of histogram relation (ablation)GloTok · ImageNet-1k reconstruction ablation (4096+12288 codebooks)PSNR: 22.30 -> 22.44Improved
histogram weight 0.01 instead of 0.1 (ablation)GloTok (40 epochs, no residual) · ImageNet-1k reconstruction ablation (4096+12288 codebooks)rFID: 1.27 -> 1.21Improved
use histogram weight 0.1 (vs none) (ablation)GloTok (40 epochs, no residual) · ImageNet-1k reconstruction ablation (4096+12288 codebooks)PSNR: 22.06 -> 21.97Hurt
swap teacher relation source: DINOv2-Kmeans vs FQGAN codebook (ablation)GloTok histogram relation · ImageNet-1k reconstructionrFID: 0.98 -> 0.95No clear effect
use GloTok tokenizer instead of FQGAN with FAR-BFAR-B · ImageNet-1k 256x256 class-conditional generationgFID: 3.38 -> 2.95Improved
use GloTok tokenizer instead of VQ-VAE tokenizer with xAR (ablation)xAR-B · ImageNet-1k 256x256 class-conditional generationgFID: 2.47 -> 2.12Improved
double codebook size to 16,384x2 (GloTok(2x)) (ablation)GloTok + xAR-B · ImageNet-1k 256x256 class-conditional generationgFID: 2.12 -> 2.00Improved
double codebook size to 16,384x2 (GloTok(2x)) (ablation)GloTok + FAR-B · ImageNet-1k 256x256 class-conditional generationgFID: 3.13 -> 2.95Improved
scale generator xAR-B to xAR-H with GloTokGloTok + xAR · ImageNet-1k 256x256 class-conditional generationgFID: 2.12 -> 1.75Improved

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