Better OCR model #10

Open

Open

Better OCR model#10

Feature

Assignees

Labels

opened

Currently using OmniParser—we should bypass and directly use a proper OCR model like PaddleOCR or EasyOCR.

Metadata

Assignees

deluzhao

Labels

Type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests