AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Moebius: 0.2B Lightweight Image Inpainting Framework with 10B-Level Performance

Kangsheng Duan, Ziyang Xu, Wenyu Liu, Xiaohu Ruan, Xiaoxin Chen, Xinggang Wang

ECCV 2026 · Image generation › Autoregressive and latent generation · approach: Generative latent space tooling · 1 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 knowledge distillation (ablation)Moebius architecture · Places2 (Test)FID: 33.42 -> 26.43Improved
add knowledge distillation (ablation)L-lambda + I-lambda Conv model · Places2 (Test)FID: 33.21 -> 24.73Improved
add fine-grained distillation, task loss and latent perceptual loss to coarse-only distillation (ablation)Moebius · Places2 (Test)FID: 74.20 -> 26.43Improved
replace standard convolutions with depthwise convolutions, no distillation (ablation)teacher-style architecture · Places2 (Test)FID: 32.75 -> 43.58Hurt
replace GLA with Local-lambda only, no distillation (ablation)teacher-style architecture · Places2 (Test)FID: 32.75 -> 37.65Hurt
replace cross-attention with Interactive-lambda only, no distillation (ablation)teacher-style architecture · Places2 (Test)FID: 32.75 -> 36.91Hurt
replace FFN with MixFFN only, no distillation (ablation)teacher-style architecture · Places2 (Test)FID: 32.75 -> 35.24Hurt
combine Local-lambda and Interactive-lambda, no distillation (ablation)teacher-style architecture · Places2 (Test)FID: 32.75 -> 33.21No clear effect
use DWConv instead of Conv with distillation (ablation)L-lambda-I-lambda-FFN · Places2 (Test)FID: 24.73 -> 25.86Mixed
add MixFFN to DWConv model with distillation (ablation)L-lambda-I-lambda · Places2 (Test)FID: 25.86 -> 26.43Mixed
isolated lightweight operators under distillation versus teacher-style with distillation (ablation)teacher-style architecture · Places2 (Test)FID: 26.81 -> 29.41Hurt
tune CFG scale to 2.5 (ablation)Moebius · Places2 (Test)FID: 10.14 -> 9.48Improved

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