SOTAVerified

Reading Order Detection

Reading order detection aims to capture the word sequence which can be naturally comprehended by human readers from visually-rich documents. It is a fundamental task for visually-rich document understanding.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–6 of 6 papers

TitleStatusHype
Infinity Parser: Layout Aware Reinforcement Learning for Scanned Document ParsingCode0
Modeling Layout Reading Order as Ordering Relations for Visually-rich Document UnderstandingCode1
Reading Order Matters: Information Extraction from Visually-rich Documents by Token Path PredictionCode1
LayerDoc: Layer-wise Extraction of Spatial Hierarchical Structure in Visually-Rich Documents—0
Relational Representation Learning in Visually-Rich Documents—0
LayoutReader: Pre-training of Text and Layout for Reading Order DetectionCode0
Show:102550
All datasetsROORReadingBank

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1LayoutLMv3-GlobalPointer (large)Segment-level F182.38—Unverified
2LayoutLMv3-GlobalPointer (base)Segment-level F168.6—Unverified
3TPP (LayoutLMv3-base)Segment-level F142.96—Unverified
4LayoutReaderSegment-level F19.44—Unverified
#ModelMetricClaimedVerifiedStatus
1LayoutReaderAverage Relative Distance (ARD)1.75—Unverified
2TPP (LayoutMask)Average Relative Distance (ARD)0.37—Unverified