AI Paper Atlas

Research map / Image generation / Image restoration

Parsing Table Structures in the Wild

Rujiao Long, Wen Wang, Nan Xue, Feiyu Gao, Zhibo Yang, Yongpan Wang, Gui-Song Xia

ICCV 2021 · Image generation › Image restoration · approach: Image harmonization and color transformation · 73 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 quadrilateral (polygon) cell regression instead of horizontal boxes (ablation)CenterNet · WTW testTEDS: 58.7 -> 70.1Improved
use quadrilateral cell regression instead of horizontal boxes (ablation)CenterNet · WTW testphysical structure F1: 73.1 -> 75.4Improved
use quadrilateral cell regression instead of horizontal boxes (ablation)CenterNet · WTW testadjacency relation F1: 84.8 -> 91.1Improved
add the cycle-pairing module (ablation)CenterNet (polygon box) · WTW testTEDS: 70.1 -> 74.3Improved
add the cycle-pairing module (ablation)CenterNet (polygon box) · WTW testphysical structure F1: 75.4 -> 78.2Improved
add the cycle-pairing module (ablation)CenterNet (polygon box) · WTW testadjacency relation F1: 91.1 -> 92.2Improved
add the pairing loss (ablation)Cycle-CenterNet · WTW testTEDS: 74.3 -> 83.3Improved
add the pairing loss (ablation)Cycle-CenterNet · WTW testphysical structure F1: 78.2 -> 78.3No clear effect
use the anchor-free CenterNet detector instead of anchor-based Cascade-RCNNcell detector baseline · WTW testTEDS: 53.2 -> 58.7Improved
use Cycle-CenterNet with pairing loss instead of vanilla CenterNettable structure parsing · WTW testTEDS: 58.7 -> 83.3Improved
use Cycle-CenterNet with pairing loss instead of vanilla CenterNettable structure parsing · WTW testadjacency relation F1: 84.8 -> 92.4Improved
train on WTW only and evaluate wired tables of ICDAR-2013 versus mixed wired and wireless tablesCycle-CenterNet · WTW testadjacency F1 (IoU 0.6): 91.7 -> 98.0Mixed

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