AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Multi-Patch Global-to-Local Transformer Architecture For Efficient Flow Matching and Diffusion Model

Quan Dao, Dimitris Metaxas

CVPR 2026 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis

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
replace isotropic DiT with multi-patch global-to-local design (k=6) (ablation)DiT-B/2 · ImageNet 256, B, 80 epochsGFLOPs: 23 -> 16.6Improved
replace isotropic DiT with multi-patch design (k=6) (ablation)DiT-B/2 + embeddings · ImageNet 256, B, 80 epochsFID: 24.52 -> 24.74No clear effect
share AdaIN module across all blocks (ablation)DiT-B/2 · ImageNet 256, B, 80 epochsFID: 34.84 -> 35.31 (Params 130.0M -> 90.3M)Hurt
use 16 class tokens instead of single class token (ablation)DiT-B/2 with shared AdaIN · ImageNet 256, B, 80 epochsFID: 35.31 -> 28.56Improved
use FNO time embedding instead of sinusoidal linear embedding (ablation)DiT-B/2 with multi-token class · ImageNet 256, B, 80 epochsFID: 28.56 -> 24.52Improved
increase number of layers in time embedding (ablation)time embedding · ImageNet 256, BFIDHurt
replace 1D grid time feature with sinusoidal time feature in FNO embedding (ablation)FNO time embedding · ImageNet 256FID: model does not convergeHurt
increase class tokens from 16 to 32 (ablation)multi-token class embedding · ImageNet 256, BFIDNo clear effect
replace ConvTranspose with Linear layer in upsample block (ablation)MPDiT · ImageNet 256, BFIDImproved
use MLP or Conv refinement after upsampling instead of one Linear layer (ablation)MPDiT upsample block · ImageNet 256, BFIDHurt
use three-level patch hierarchy {8,4,2} instead of two-level {4,2} (ablation)MPDiT-XL · ImageNet 512FID: FID 2.47 -> 3.13 with cfg; GFLOPs 228.4 -> 138.2Mixed
use MPDiT-XL instead of SiT-XL/2MPDiT-XL · ImageNet 256, 80 epochsFID: 18.04 -> 9.92, GFLOPs 118.66 -> 59.30Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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