Our new X account is live! Follow @wizwand_team for updates

Home/Benchmarks

Object Grounding on Refclef (val)

58.9Accuracy

PositionOCR

Updated 4d ago

Evaluation Results

Method	Links
PositionOCR 2026.02		58.9

SOTA Paper

PositionOCR

PositionOCR: Augmenting Positional Awareness in Multi-Modal Models via Hybrid Specialist Integration

Dataset

RefCLEF

Follow for update

@wizwand_team Discord

Related Benchmarks

Document Information Extraction on DeepForm Object Grounding on Refcoco+ (TestA)Information Extraction on POIE Text Grounding on DocLocal4K (test)Text Spotting on ICDAR 2015

© 2026 wizwand

Blog Contact Changelog

Privacy Policy Terms of Service FAQs