ViSceT5 Collection Vietnamese scene-text VQA: ViT5 + QA-ViT + Consformer + OCR, final models. • 3 items • Updated about 22 hours ago