AI Paper Atlas

Research map / Image generation / Image restoration

TAPE: Task-Agnostic Prior Embedding for Image Restoration

Lin Liu, Lingxi Xie, Xiaopeng Zhang, Shanxin Yuan, Xiangyu Chen, Wengang Zhou, Houqiang Li, Qi Tian

ECCV 2022 · Image generation › Image restoration · approach: Priors for restoration networks · 69 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 task-agnostic pre-training (ablation)TAPE-Net · Rain200LPSNR: 31.72 -> 33.17Improved
add task-agnostic pre-training (ablation)TAPE-Net · SIDDPSNR: 37.41 -> 37.90Improved
add task-agnostic pre-training then fine-tune on unseen task (ablation)TAPE-Net · Snow100KPSNR: 25.42 -> 26.33Improved
add task-agnostic pre-training then fine-tune on unseen task (ablation)TAPE-Net · ISTDPSNR: 26.28 -> 26.57Improved
add pre-training for IPT versus TAPE on unseen tasksIPT vs TAPE-Net · Snow100K/ISTD/REDSPSNR gain: IPT boosts 0.07dB, 0.19dB and 0.08dB; TAPE improves 0.91dB, 0.29dB and 0.48dBImproved
widen synthetic Gaussian noise range in pre-training from [5,20] to [1,50] (ablation)TAPE-Net · SIDDPSNR: 37.60 -> 37.72Improved
pre-train with 5 datasets instead of 3 (ablation)TAPE-Net · Snow100KPSNR: 26.17 -> 26.33Improved
add more tasks to pre-training (ablation)TAPE-Net · pre-training-known tasks (Raindrop800, Rain200L, TIP2018)PSNR: maximum PSNR difference is 0.04 dBNo clear effect
add pre-training of network theta (ablation)TAPE-Swin · Rain200HPSNR: 25.96 -> 26.18Improved
add pre-training of network phi (ablation)TAPE-Swin · Rain200HPSNR: 26.15 -> 26.18Improved
add pre-training of theta and PLM (ablation)TAPE-Swin · Rain200HPSNR: 25.73 -> 26.18Improved
add pixel-wise contrastive loss in pre-training (ablation)TAPE-Net · Rain200LPSNR: decrease by 0.12dB/0.001 without the lossImproved

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