search

ultralytics / yolov5

YOLOv5 🚀 in PyTorch > ONNX > CoreML > TFLite
https://docs.ultralytics.com
GNU Affero General Public License v3.0
51.22k stars 16.44k forks source link

Any trained data for QR codes and barcodes? #5191

Closed so-o closed 3 years ago

so-o commented 3 years ago

❔Question

Question is in the title.

I found trained data for QR codes for yolov4, not for barcodes. I need to convince some people before we start generating our own data. The trained data set I have for yolov4 doesn't work when 2 QR codes are close to each other.

Additional context

github-actions[bot] commented 3 years ago

👋 Hello @so-o, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide screenshots and minimum viable code to reproduce your issue, otherwise we can not help you.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset images, training logs, screenshots, and a public link to online W&B logging if available.

For business inquiries or professional support requests please visit https://ultralytics.com or email Glenn Jocher at glenn.jocher@ultralytics.com.

Requirements

Python>=3.6.0 with all requirements.txt installed including PyTorch>=1.7. To get started:

$ git clone https://github.com/ultralytics/yolov5
$ cd yolov5
$ pip install -r requirements.txt

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

Status

CI CPU testing

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training (train.py), validation (val.py), inference (detect.py) and export (export.py) on MacOS, Windows, and Ubuntu every 24 hours and on every commit.

glenn-jocher commented 3 years ago

@so-o we don't have a QR dataset ready to go, but if you find or create one it would be super useful if you could submit a PR to add it to our autodownload datasets here to allow other users to get started more easily in the future: https://github.com/ultralytics/yolov5/tree/master/data

About improving your training results, see Tips for Best Training Results tutorial below. It was written for YOLOv5, but applies broadly to Vision AI in general, so should also help with YOLOv4 training.

YOLOv5 Tutorials

softmatic commented 3 years ago

@so-o, do you want to do actual barcode recognition or just tell if there's a barcode in the image? We do barcodes for a living (softmatic.com) and can easily mass create barcodes for synthetic data.

glenn-jocher commented 3 years ago

@softmatic @so-o one important point to remember is that the background/surroundings of the QR/bar codes are just as important as the codes themselves, and this may not be easily simulated. i.e. if you create a synthetic dataset of codes on white/solid backgrounds this will probably fail (poor generalization, excessive FPs) when deployed in the real world. The train, val and deployed image spaces must all be subsets of a common image space for best results:

dataset
softmatic commented 3 years ago

Yeah, it's not trivial. I've been looking into generating synthetic data for another project, there's a very extensive paper that reviews the various approaches and pitfalls: arxiv

github-actions[bot] commented 3 years ago

👋 Hello, this issue has been automatically marked as stale because it has not had recent activity. Please note it will be closed if no further activity occurs.

Access additional YOLOv5 🚀 resources:

Access additional Ultralytics ⚡ resources:

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐!