I need to annotate documents for the OCR task. I wanted to setup a following workflow:

1. Autolabel data using the EasyOCR backend.
2. Interactively label missing spots (because EasyOCR often doesn't detect some words) using the Tesseract backend.

But I faced issue that I cannot connect these two backends at least via UI.
Is it possible to connect multiple backends to the one project natively?
If not, maybe someone can suggest some workarounds that can help me implement my workflow?