From Observed Model Assets to Tiny Local Document Intelligence
Year: 2026
Type: preprint
Abstract: A methods note showing how a Mercury observed-model archive can be turned into a teacher inventory for tiny local document-intelligence packs, with explicit OCR boundaries, privacy-safe layout ghosts, 50MB/100MB size gates, and app-side release gates.
Author: Chen, Ho Yiing (norika)
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7165016136
Cites:
Cited by:
Related to: