Receipt Detection Dataset for YOLO

Description

This dataset contains annotated receipt images prepared in YOLO format for object detection tasks. It combines multiple public sources, with duplicate removal, normalization, and standardized train/validation/test splits to ensure quality and reproducibility. The dataset provides one of the few open resources for training and benchmarking YOLO-based models in receipt detection, supporting applications in computer vision, document analysis, and financial AI systems such as automated expense tracking and digital archiving.

Steps to reproduce

Download the dataset and unzip the files. Each image has a corresponding YOLO .txt annotation file with normalized coordinates. Train/validation/test splits are provided. The dataset can be directly used with YOLOv5, YOLOv7, or YOLOv8 for receipt detection tasks.

Institutions

• Multimedia University

  Melaka, Malacca

Categories

Licence