Patent · US Active

Unified pretraining framework for document understanding

US12333845B2 · kind B2 · utility

0Cited by
1References
15Claims
0Family size

Assignee

Inventors

Key dates

Filing dateNov 16, 2021
Grant dateJun 17, 2025
Priority date
Expiry dateMay 31, 2043

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06V30/147
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

The technology described includes methods for pretraining a document encoder model based on multimodal self cross-attention. One method includes receiving image data that encodes a set of pretraining documents. A set of sentences is extracted from the image data. A bounding box for each sentence is generated. For each sentence, a set of predicted features is generated by using an encoder machine-learning model. The encoder model performs cross-attention between a set of masked-textual features for the sentence and a set of masked-visual features for the sentence. The set of masked-textual features is based on a masking function and the sentence. The set of masked-visual features is based on the masking function and the corresponding bounding box. A document-encoder model is pretrained based on the set of predicted features for each sentence and pretraining tasks. The pretraining tasks includes masked sentence modeling, visual contrastive learning, or visual-language alignment.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.