AI Paper Atlas

Research map / Optimization / Optimizers

Proximal Alternating Direction Network: A Globally Converged Deep Unrolling Framework

Risheng Liu, Xin Fan, Shichao Cheng, Xiangyu Wang, Zhongxuan Luo

AAAI 2018 · Optimization › Optimizers · approach: Constrained and non-backprop training · 28 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 PADNet with a learned prior and explicit task fidelityNon-blind deconvolution · Levin benchmarkPSNR: 32.74 -> 33.41Improved
Use PADNet with a learned prior and explicit task fidelityNon-blind deconvolution · Sun benchmarkPSNR: 32.44 -> 32.69Improved
Use implicit PADNet to learn priors from training dataNon-blind deconvolution · Sun benchmarkPSNR: 32.44 -> 32.71Improved
Use PADNet for haze removalSingle-image haze removal · Fattal benchmarkPSNR: 27.11 -> 28.47Improved
Use PADNet for haze removalSingle-image haze removal · Fattal benchmarkTime (s): 3.73 -> 2.74Improved
Replace conventional optimization iterations with learned PADNet propagationGradient-domain deconvolution · Levin benchmarkNumber of iterations: 29 -> 5Improved
Use the learned implicit prior instead of handcrafted ℓ1 or ℓ0 priorsGradient-domain deconvolution · Levin benchmarkRelative reconstruction error: 1.3146 -> 0.4291; 1.7114 -> 0.4291Improved
Use the learned implicit prior rather than the explicit ℓ0.8 priorGradient-domain deconvolution · Levin benchmarkRelative reconstruction error: 0.4140 -> 0.4291Hurt
Choose explicit PADNet with ℓ0.8 for standard images and implicit PADNet for challenging real-world imagesNon-blind deconvolution · Levin and Sun benchmarksPSNR: 33.41 vs 33.37; 32.69 vs 32.71Mixed

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