Can I get the reading order，if I use Layout predict Model?

[CourageKeenTriumph]

Can I get the reading order，if I use Layout predict Model?

[GreatV]

No, the LayoutXLM model does not explicitly provide reading order as part of its output. LayoutXLM is designed for multi-modal pre-training for visually-rich document understanding, and it includes tasks such as Semantic Entity Recognition (SER) and Relation Extraction (RE). These tasks help classify text elements and relationships but do not inherently define a strict reading order.

If you need to determine the reading order, you may need to:

1. Use text block coordinates extracted from LayoutXLM and apply a heuristic approach (such as sorting by the Y-axis first, then X-axis).
2. Leverage OCR output that typically contains reading order information.
3. Integrate a document layout analysis technique to infer reading sequences.

Would you like recommendations on methods to infer reading order from layout models?

---

Response generated by 🤖 feifei-bot | chatgpt-4o-latest

⚠️ This is an AI-generated response. Please verify any critical information.