AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Total Deep Variation: A Stable Regularization Method for Inverse Problems

Erich Kobler, Alexander Effland, Karl K. Kunisch, Thomas Pock

TPAMI 2021 · Adversarial robustness and security › Certified and empirical robustness · approach: Decision-boundary geometry of robustness · 59 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
use identity potential psi(x)=x instead of ln cosh(x) (ablation)TDV_3^3 · BSDS68 gray denoising sigma=25PSNR: 29.3596 -> 29.3722Improved
use identity potential psi(x)=x instead of 1/2 log(1+x^2) (ablation)TDV_3^3 · BSDS68 gray denoising sigma=25PSNR: 29.3662 -> 29.3722Improved
train TDV individually per noise level instead of one model trained for sigma=25TDV_3^3 · Urban100 gray denoising sigma=50PSNR: 26.94 -> 27.04Improved
train TDV individually per noise level instead of one model trained for sigma=25TDV_3^3 · Urban100 gray denoising sigma=15PSNR: 32.66 -> 32.91Improved
train TDV individually per noise level instead of one model trained for sigma=25TDV_3^3 · BSDS68 gray denoising sigma=25PSNR: 29.37 -> 29.37No clear effect
increase number of blocks and scales of the TDV regularizer (ablation)TDV · BSDS68 gray denoising sigma=25expected PSNRImproved
increase depth S beyond 10 (ablation)TDV · BSDS68 gray denoising sigma=25expected PSNRNo clear effect
stop gradient flow at optimal stopping time T*=0.172 rather than running longer (ablation)TDV_3^3 · BSDS68 gray denoising sigma=25PSNR: T*=0.172Improved
use TDV_3^3 instead of FOCNetTDV · gray denoising parameter counttrainable parameters: 53,513,120 -> 387,394Improved
use TDV_{3,25}^3 instead of FFDNetTDV · Kodak24 color denoising sigma=50PSNR: 28.98 -> 29.44Improved
use TDV_{3,25}^3 instead of FFDNetTDV · McMaster color denoising sigma=15PSNR: 34.66 -> 34.55Hurt
use TDV_3^3 instead of OISR-LF-sTDV · Set14 super-resolution x2PSNR: 33.62 -> 33.35Hurt

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