Bank Name Region Detection Dataset

Bank Name Region Detection Dataset A publicly available object detection dataset for Bank Name Region Localization in financial documents. The dataset is annotated in the COCO Object Detection format and is intended for training, fine-tuning and benchmarking document understanding and OCR-aware object detection models. Dataset Statistics Split Images Train 5,580 Validation 264 Test 132 Total 5,976 Classes ID Class… See the full description on the dataset page:

Source: Hugging Face Hub (Srijan-Chakraborty/Bank-Name-Region-Detection-Dataset). Metadata imported from the dataset’s Hub tags.