Skip to content

YOLO.md

Latest commit

 

History

History
37 lines (27 loc) · 1.64 KB

YOLO.md

File metadata and controls

37 lines (27 loc) · 1.64 KB
Raw

Preparing YOLO datasets

Make sure all datasets are located in datasets folder. Currently we are using these datasets:

  1. taco-trash-dataset
  2. Plastic - Paper - Garbage Bag Synthetic Images
  3. WasteInsight: Dataset for Detection and Volume Estimation of Municipal Solid Waste

Given varying dataset formatting and initial COCO format first we need to convert those sets into YOLO format:

1. taco-trash-dataset

From the root directory run:

uv run src/scripts/preprocess_taco.py

uv run coco_to_yolo datasets/_taco-trash-dataset_processed datasets_yolo/ --test_ratio 0.15 --val_ratio 0.1

mv datasets_yolo/converted datasets_yolo/taco-trash-dataset_yolo && mv datasets_yolo/converted.yaml datasets_yolo/taco-trash-dataset_yolo/classes.yaml

2. Synthetic Bags

uv run src/scripts/preprocess_synthetic_bags.py
uv run coco_to_yolo datasets/_synthetic-bags_processed datasets_yolo/ --test_ratio 0.15 --val_ratio 0.1
mv datasets_yolo/converted datasets_yolo/synthetic-bags_yolo && mv datasets_yolo/converted.yaml datasets_yolo/synthetic-bags_yolo/classes.yaml

3. WasteInsight

From the root directory run:

uv run src/scripts/preprocess_wasteinsight.py
uv run coco_to_yolo datasets/_wasteinsight_processed datasets_yolo/ --test_ratio 0.15 --val_ratio 0.1
mv datasets_yolo/converted datasets_yolo/wasteinsight_yolo && mv datasets_yolo/converted.yaml datasets_yolo/wasteinsight_yolo/classes.yaml