AI Paper Atlas

Research map / Image generation / Image restoration

HiNet: Deep Image Hiding by Invertible Network

Junpeng Jing, Xin Deng, Mai Xu, Jianyi Wang, Zhenyu Guan

ICCV 2021 · Image generation › Image restoration · approach: Inverse imaging with learned priors · 408 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 wavelet transform (hide in wavelet domain) (ablation)HiNet · cover/stego pairPSNR (dB): 40.71 -> 44.23Improved
add wavelet transform (hide in wavelet domain) (ablation)HiNet · secret/recovery pairPSNR (dB): 44.22 -> 48.52Improved
add low-frequency wavelet loss Lfreq (ablation)HiNet · SRNet detectiondetection rate (%): 75.42 -> 55.86Improved
add low-frequency wavelet loss Lfreq (ablation)HiNet · cover/stego pairPSNR (dB): 44.23 -> 46.52Improved
add low-frequency wavelet loss Lfreq (ablation)HiNet · secret/recovery pairPSNR (dB): 48.52 -> 46.98Hurt
use dense blocks for rho, phi, eta instead of convolutional or residual blocks (ablation)HiNet · cover/stego PSNRPSNR (dB)Improved
set lambda_freq to 10 (ablation)HiNet · PSNR, SSIM and detection ratePSNR (dB), SSIM, detection rateMixed
use invertible network with shared concealing/revealing parameters instead of two separate networksBaluja · DIV2K cover/stegoPSNR (dB): 36.77 -> 48.99Improved
use invertible network with shared parameters instead of Weng et al.Weng et al. · DIV2K secret/recoveryPSNR (dB): 38.93 -> 52.86Improved
use HiNet instead of Weng et al. for steganalysis resistanceWeng et al. · SRNet detectiondetection accuracy (%): 75.03 -> 55.86Improved

Later papers that compared with this method

21 later accepted papers report HiNet as a baseline in their results tables; 19 of them report a better number on most of the results they share with it.

13 more not listed.

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