AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

LaWa: Using Latent Space for In-Generation Image Watermarking

Ahmad Rezaei, Mohammad Akbari, Saeed Ranjbar Alvar, Arezou Fatemi, Yong Zhang

ECCV 2024 · Image generation › Autoregressive and latent generation · approach: Robust diffusion watermarking · 44 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
embed watermark in latent space of frozen decoder (in-generation) instead of post-generation encoder-decoder or optimization methodsLaWa · AI-generated images, 32-bitaverage bit accuracy: 0.92 -> 0.99 vs SSL/RivaGanImproved
in-generation embedding instead of post-generation processingLaWa · AI-generated imagesembedding time (ms): SSL 870 -> LaWa 1Improved
use LaWa as post-generation instead of in-generationLaWa · AI-generated imagesaverage bit accuracy: 0.99 -> 0.99 (32-bit); 0.97 -> 0.97 (48-bit)No clear effect
LaWa vs FNNS under 10% random crop at 48 bitsLaWa · AI-generated images, 48-bitbit accuracy: FNNS 0.96 vs LaWa 0.91Hurt
increase message length from 32 to 128 bits (ablation)LaWa · AI-generated imagesbit accuracy (combined attack): 0.97 -> 0.87Hurt
increase quality-robustness weight lambda from 1.0 to 10.0 (ablation)LaWa · AI-generated imagesbit accuracy on combined attack / PSNR: bit acc 0.65 -> 0.99; PSNR 34.76 -> 28.47Mixed
add Critic loss to MSE and LPIPS losses (ablation)LaWa · AI-generated imagesPSNR/SSIM: 29.47/0.47 -> 33.52/0.86Improved
add LPIPS loss to MSE plus Critic (ablation)LaWa · AI-generated imagesLPIPS/SIFID: 0.10/0.05 -> 0.04/0.02Improved
add LPIPS loss to MSE plus Critic (ablation)LaWa · AI-generated imagesbit accuracy on combined attack: 0.99 -> 0.94Hurt
raise image MSE loss weight lambda_I above 0.1 (ablation)LaWa · AI-generated imagesLPIPS/SIFID and robustness: bit acc comb. 0.94 -> 0.84 at lambda_I=1.0Hurt
LaWa vs Stable Signature against autoencoder removal attack (KL-VAE f8)LaWa · learning-based attacksbit accuracy: 55% -> 85%Improved

Later papers that compared with this method

6 later accepted papers report Lawa(48) as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.

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