AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

DiT-IC: Aligned Diffusion Transformer for Efficient Image Compression

Junqi Shi, Ming Lu, Xingchen Li, Anle Ke, Ruiqi Zhang, Zhan Ma

CVPR 2026 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis · 4 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 alignment strategies (variance-guided flow, self-distillation, latent condition) to DiT baseline (ablation)DiT-IC · Avg BD-rate LPIPSBD-rate (%): -65.00 -> -83.65Improved
add alignment strategies to DiT baseline (ablation)DiT-IC · Avg BD-rate DISTSBD-rate (%): -67.19 -> -87.88Improved
initialize from pretrained SANA instead of training DiT from scratch (ablation)DiT-IC · ablationBD-rate DISTS (%): 32.45% -> 0%Improved
use LoRA rank 32/64 instead of full finetuning (ablation)DiT-IC · ablationBD-rate DISTS (%): 8.05% -> 0%Improved
use LoRA rank 32/64 instead of rank 16/16 (ablation)DiT-IC · ablationBD-rate DISTS (%): 13.92% -> 0%Improved
add adversarial loss Ladv (ablation)DiT-IC · ablationBD-rate PSNR (%): -37.10% -> 0%Hurt
add adversarial loss Ladv (ablation)DiT-IC · ablationBD-rate DISTS (%): -1.80% -> 0%Improved
add DISTS term to distortion loss (ablation)DiT-IC · ablationBD-rate DISTS (%): 5.69% -> 0%Improved
add DISTS term to distortion loss (ablation)DiT-IC · ablationBD-rate LPIPS (%): -1.83% -> 0%Hurt
use DiT in 32x latent space instead of U-Net diffusion (StableCodec)DiT-IC · 1024x1024 FP32diffusion latency (s): 0.11 -> 0.055Improved
use DiT-IC at 4096x4096DiT-IC · 4096x4096 FP32total latency (s): 85 -> 3.3Improved
use DiT-IC VAE vs SD-VAE at 1024x1024DiT-IC · 1024x1024 FP32VAE latency (s): 0.19 -> 0.21Hurt

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