Srijan-Chakraborty/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
1	bank_name

---

Annotation Format

COCO Object Detection

Each annotation contains:

• Bounding Box
• Category ID
• Area
• Crowd Flag
• Segmentation

---

Folder Structure

train/
    images/
    annotations.coco.json

valid/
    images/
    annotations.coco.json

test/
    images/
    annotations.coco.json

---

Applications

This dataset can be used for

• OCR Pipelines
• Document AI
• Invoice Understanding
• Financial Document Processing
• Information Extraction
• Object Detection
• Vision Language Models
• Document Layout Analysis

---

Supported Frameworks

• YOLO (v5/v8/v11)
• RT-DETR
• Detectron2
• MMDetection
• Faster R-CNN
• SSD
• EfficientDet
• Any COCO-compatible framework

---

License

This dataset is released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license.

You are free to

• Share
• Modify
• Redistribute
• Use commercially

provided proper attribution is given.

---

Author

Srijan Chakraborty

GitHub

https://github.com/SrijanChakraborty2003

Hugging Face

https://huggingface.co/Srijan-Chakraborty

---

Contact

For bug reports, feature requests, or suggestions, please open a GitHub Issue or Discussion.

If you find this dataset useful, consider following my GitHub and Hugging Face profiles for future datasets and updates.

---

Citation

@dataset{chakraborty2026bankname,
  author    = {Srijan Chakraborty},
  title     = {Bank Name Region Detection Dataset},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/Srijan-Chakraborty/Bank-Name-Region-Detection-Dataset}
}

---

Acknowledgements

This dataset was manually curated and verified for public release to support research and development in OCR, Document AI and object detection.