145 labeled images of Optical Mark Recognition (OMR) bubble sheets. It is designed to feature real-world edge cases for computer vision tasks.
Use Cases
- Train a computer vision model to detect and read marked bubbles on OMR sheets using the 145 labeled images.
- Develop a document processing pipeline for OMR bubble sheets that handles real-world edge cases present in the image collection.
- Benchmark the robustness of Optical Mark Recognition algorithms against challenging, non-ideal image conditions.
Strengths
- Contains 145 labeled images, providing a focused collection for model training and testing.
- Specifically features real-world edge cases, which are valuable for testing algorithm robustness.
Limitations
- Small sample size of 145 images may limit the ability to train large, deep learning models from scratch.
- Lack of detailed metadata (e.g., column descriptions, file formats) limits understanding of the data structure.