Plastic Paper Garbage Bag Synthetic Images
Synthetic dataset composed of 30,000 object images (plastic bags, paper and waste) with COCO annotations for classification and segmentation. The images are grouped by classes and combined into a folder with multi-class annotations.
Approximately 30,000 synthetic images annotated in COCO format, plastic categories, paper, waste bags
CC BY-SA 4.0
Description
The dataset Plastic Paper Garbage Bag Synthetic Images contains approximately 30,000 synthetic images depicting plastic bags, paper bags, and trash. The images are annotated in COCO format, allowing classification, detection and segmentation tasks. This dataset is particularly suitable for applications in the field of recycling and waste management.
What is this dataset for?
- Train CNN models for the classification and segmentation of objects in an environmental context
- Develop object recognition systems for waste management and recycling
- Testing computer vision algorithms on various synthetic images
Can it be enriched or improved?
Yes, this dataset can be completed with real images to reinforce robustness, annotated more finely for segmentation, or remixed with synthetic variants via image augmentation.
🔎 In summary
🧠 Recommended for
- Computer vision researchers
- Environmental projects
- Deep learning engineers
🔧 Compatible tools
- COCO API
- TensorFlow
- PyTorch
- Detectron2
💡 Tip
Combine this dataset with real images to improve the generalization of models.
Frequently Asked Questions
Is this dataset only synthetic or does it contain real images?
It is completely synthetic, created from royalty-free images, and generated for diversity.
What annotation format is used?
The COCO format is used, suitable for classification, detection and segmentation.
Can I use this dataset for a commercial project?
Yes, provided that you respect the CC BY-SA 4.0 license, including like-sharing.
