Codesota · Benchmark · lam(line-level)Home/Leaderboards/Vision & Documents/Document OCR/lam(line-level)
Unknown

lam(line-level).

lam(line-level) is a state-of-the-art machine learning benchmark indexed on Codesota. This page tracks published model results, top scores per metric, and the SOTA timeline for lam(line-level).

Paper ↗Leaderboard ↓
§ 01 · SOTA history

Year over year.

§ 02 · Leaderboard

Results by metric.

Found a wrong score or missing run?
Use row edits to send a sourced correction into moderation.
Add / edit result ↗Report issue ↗

Test Wer

Test Wer is the reported evaluation metric for lam(line-level). Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for Test Werverifiedpapervendorcommunityunverified
RankModelTrustScoreYearLinksFix
01GFCN
From paper: Recurrence-free unconstrained handwritten text recognition using gated fully convolutional network
verified18.52020Paper ↗Code ↗Looks wrong?
02TrOCR
From paper: TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
verified11.62021Paper ↗Code ↗Looks wrong?
03OrigamiNet-12
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified11.22020Paper ↗Code ↗Looks wrong?
04OrigamiNet-18
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified11.12020Paper ↗Code ↗Looks wrong?
05OrigamiNet-24
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified112020Paper ↗Code ↗Looks wrong?
06HTR-VT
From paper: HTR-VT: Handwritten Text Recognition with Vision Transformer
verified7.402024Paper ↗Code ↗Looks wrong?
07HTR-ConvText
From paper: HTR-ConvText: Leveraging Convolution and Textual Information for Handwritten Text Recognition (Table 3). New SOTA on LAM line-level.
verified7.002024Paper ↗Looks wrong?

Test Cer

Test Cer is the reported evaluation metric for lam(line-level). Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for Test Cerverifiedpapervendorcommunityunverified
RankModelTrustScoreYearLinksFix
01GFCN
From paper: Recurrence-free unconstrained handwritten text recognition using gated fully convolutional network
verified5.202020Paper ↗Code ↗Looks wrong?
02TrOCR
From paper: TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
verified3.602021Paper ↗Code ↗Looks wrong?
03OrigamiNet-12
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified3.102020Paper ↗Code ↗Looks wrong?
04OrigamiNet-18
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified3.102020Paper ↗Code ↗Looks wrong?
05OrigamiNet-24
From paper: OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
verified3.002020Paper ↗Code ↗Looks wrong?
06HTR-VT
From paper: HTR-VT: Handwritten Text Recognition with Vision Transformer
verified2.802024Paper ↗Code ↗Looks wrong?
07HTR-ConvText
From paper: HTR-ConvText: Leveraging Convolution and Textual Information for Handwritten Text Recognition (Table 3). New SOTA on LAM line-level, surpassing HTR-VT (2.8% CER).
verified2.702024Paper ↗Looks wrong?
§ 04 · Submit a result

Add to the leaderboard.

← Back to Document OCR