AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Beyond Image Prior: Embedding Noise Prior into Latent Space of Conditional Denoising Transformer

Yuanfei Huang, Hua Huang

IJCV 2025 · Image generation › Autoregressive and latent generation · approach: Hierarchical VQ-VAE generation · 9 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 CondSA block with zero-vector prior (ablation)Restormer · SIDD ValidationPSNR: 39.97 -> 39.96No clear effect
embed estimated noise prior by naive concatenation in CondSA (ablation)Restormer · SIDD ValidationPSNR: 39.96 -> 40.09Improved
add linear fusion module (LFM) (ablation)Condformer · SIDD ValidationPSNR: 40.09 -> 40.21Improved
use ground-truth noise prior instead of estimated (ablation)Condformer · SIDD ValidationPSNR: 40.21 -> 40.23Improved
embed prior into Q, K and V instead of Q, K only (ablation)Condformer · SIDD ValidationPSNR: 40.23 -> 40.19Hurt
add local smoothness criterion to patch sampling (ablation)LoNPE algorithm · Urban100RMSE: 0.125 -> 0.021Improved
add local smoothness criterion to patch sampling (ablation)LoNPE algorithm · SIDD-MediumCV: 0.213 -> 0.048Improved
increase patch size from 16x16 to 32x32 (ablation)LoNPE algorithm · Urban100RMSE: 0.020 -> 0.045Hurt
increase sampling ratio from 10% to 20% (ablation)LoNPE algorithm · Urban100RMSE: 0.020 -> 0.031Hurt
condition denoiser on noise prior (Condformer vs Restormer re-trained on P-G)Restormer · CBSD68 (sigma_s=0.3)PSNR: 28.01 -> 28.14Improved
use Condformer vs Restormer on Gaussian noise (sigma_s=0)Restormer trained on Gaussian · Kodak24 (sigma_s=0)PSNR: 32.82 -> 32.81Hurt
condition denoiser on noise priorRestormer · SIDD ValidationPSNR: 40.02 -> 40.21Improved

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