WeiChihChern/Copy-Paste-Semantic-Segmentation
Simple Copy and Paste Implementation for Semantic Segmentation (Data Augmentation)
This tool helps machine learning engineers expand their image datasets for semantic segmentation tasks. By taking an original image and its corresponding mask (where each pixel intensity indicates a class), it generates new, augmented images. These new images creatively 'copy and paste' objects from existing images onto different backgrounds, rotating and scaling them to create diverse training examples. The end-user is a machine learning practitioner training models for image analysis.
No commits in the last 6 months.
Use this if you need to artificially increase the size and diversity of your image datasets for semantic segmentation, especially when you have limited real-world annotated images.
Not ideal if you require instance-level segmentation augmentation where individual objects are copied and pasted independently, as this focuses on copying entire segments.
Stars
7
Forks
—
Language
Jupyter Notebook
License
MIT
Category
Last pushed
Mar 15, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/WeiChihChern/Copy-Paste-Semantic-Segmentation"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
TorchIO-project/torchio
Medical imaging processing for AI applications.
aleju/imgaug
Image augmentation for machine learning experiments.
makcedward/nlpaug
Data augmentation for NLP
mdbloice/Augmentor
Image augmentation library in Python for machine learning.
BloodAxe/pytorch-toolbelt
PyTorch extensions for fast R&D prototyping and Kaggle farming