DocVAL transfers spatial reasoning via validated CoT distillation from large teachers to compact student VLMs, delivering up to 6-7 ANLS gains and strong mAP localization on document VQA benchmarks.
Icdar2019 competi- tion on scanned receipt ocr and information extraction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DocVAL: Validated Chain-of-Thought Distillation for Grounded Document VQA
DocVAL transfers spatial reasoning via validated CoT distillation from large teachers to compact student VLMs, delivering up to 6-7 ANLS gains and strong mAP localization on document VQA benchmarks.