Dear OpenCV Development Team,
Recently, I submitted my proposal about Developing an Advanced QR/Barcode/ArUco Detector into GSoC. However, I have a couple of questions:
1. Are there any preferred datasets for training, or should we collect and augment our own?
2. Are there any constraints regarding API changes or backward compatibility with existing OpenCV functionalities?
3. Are there any specific benchmarks or tests that the model must pass before being considered a success?
Any reply would be appreciated, thank you!
Matthew Wong