Papers › StrucTexTv2: Masked Visual-Textual Prediction for Document Image Pre-training

StrucTexTv2: Masked Visual-Textual Prediction for Document Image Pre-training

1 Mar 2023arXiv:2303.00289archive 2025-07-28

Yuechen Yu, Yulin Li, Chengquan Zhang, Xiaoqiang Zhang, Zengyuan Guo, Xiameng Qin, Kun Yao, Junyu Han, Errui Ding, Jingdong Wang

In this paper, we present StrucTexTv2, an effective document image pre-training framework, by performing masked visual-textual prediction. It consists of two self-supervised pre-training tasks: masked image modeling and masked language modeling, based on text region-level image masking. The proposed method randomly masks some image regions according to the bounding box coordinates of text words. The objectives of our pre-training tasks are reconstructing the pixels of masked image regions and the corresponding masked tokens simultaneously. Hence the pre-trained encoder can capture more textual semantics in comparison to the masked image modeling that usually predicts the masked image patches. Compared to the masked multi-modal modeling methods for document image understanding that rely on both the image and text modalities, StrucTexTv2 models image-only input and potentially deals with more application scenarios free from OCR pre-processing. Extensive experiments on mainstream benchmarks of document image understanding demonstrate the effectiveness of StrucTexTv2. It achieves competitive or even new state-of-the-art performance in various downstream tasks such as image classification, layout analysis, table structure recognition, document OCR, and information extraction under the end-to-end scenario.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

PaddlePaddle/VIMER officialpaddle report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Document Image ClassificationImage ClassificationLanguage ModelingLanguage ModellingMasked Language ModelingOptical Character Recognition (OCR)Semantic entity labelingimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Document Image Classification RVL-CDIP StrucTexTv2 (large) Accuracy 94.62% #14 of 31 Archive leaderboard report
Document Image Classification RVL-CDIP StrucTexTv2 (large) Parameters 238M #14 of 31 Archive leaderboard report
Document Image Classification RVL-CDIP StrucTexTv2 (small) Accuracy 93.4% #18 of 31 Archive leaderboard report
Document Image Classification RVL-CDIP StrucTexTv2 (small) Parameters 28M #18 of 31 Archive leaderboard report
Semantic entity labeling FUNSD StrucTexTv2 (large) F1 91.82 #7 of 15 Archive leaderboard report
Semantic entity labeling FUNSD StrucTexTv2 (small) F1 89.23 #9 of 15 Archive leaderboard report
Table Recognition WTW StrucTexTv2 (small) F1 78.9% #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections