Cleanlab/object-detection-tutorial

1---2license: mit3task_categories:4- object-detection5language:6- en7tags:8- computer-vision9- cleanlab10- data-centric-ai11- bounding-boxes12pretty_name: Object Detection Tutorial Dataset13size_categories:14- n<1K15---16 17# Object Detection Tutorial Dataset18 19## Dataset Description20 21This dataset contains object detection annotations and predictions used in the cleanlab tutorial: [Object Detection](https://docs.cleanlab.ai/stable/tutorials/object_detection.html).22 23The dataset demonstrates how to use cleanlab to identify and correct label issues in object detection datasets, where labels consist of bounding boxes around objects in images.24 25### Dataset Summary26 27- **Total Examples**: 118 images with bounding box annotations28- **Task**: Object detection with bounding boxes29- **Files**:30  - `labels.pkl`: Ground truth bounding box labels31  - `predictions.pkl`: Model predictions for bounding boxes32  - `example_images.zip`: Sample images for object detection33 34### Dataset Structure35 36```python37from huggingface_hub import hf_hub_download38import pickle39import zipfile40 41# Download labels42labels_path = hf_hub_download('Cleanlab/object-detection-tutorial', 'labels.pkl')43with open(labels_path, 'rb') as f:44    labels = pickle.load(f)45 46# Download predictions47predictions_path = hf_hub_download('Cleanlab/object-detection-tutorial', 'predictions.pkl')48with open(predictions_path, 'rb') as f:49    predictions = pickle.load(f)50 51# Download and extract images52images_path = hf_hub_download('Cleanlab/object-detection-tutorial', 'example_images.zip')53with zipfile.ZipFile(images_path, 'r') as zip_ref:54    zip_ref.extractall('example_images/')55```56 57### Data Format58 59- **labels.pkl**: Dictionary containing ground truth bounding boxes in format `[x_min, y_min, x_max, y_max, class_id]`60- **predictions.pkl**: Dictionary containing predicted bounding boxes with confidence scores61- **example_images.zip**: Compressed folder containing image files62 63## Dataset Creation64 65This dataset was created for educational purposes to demonstrate cleanlab's capabilities for detecting issues in object detection datasets, such as:66- Incorrectly labeled bounding boxes67- Missing annotations68- Poor quality predictions69- Annotation inconsistencies70 71## Uses72 73### Primary Use Case74 75This dataset is designed for:761. Learning data-centric AI techniques for object detection772. Demonstrating cleanlab's object detection issue detection783. Teaching proper annotation quality assessment workflows79 80### Example Usage81 82```python83from huggingface_hub import hf_hub_download84import pickle85from cleanlab.object_detection.summary import object_detection_health_summary86 87# Download files88labels_path = hf_hub_download('Cleanlab/object-detection-tutorial', 'labels.pkl')89predictions_path = hf_hub_download('Cleanlab/object-detection-tutorial', 'predictions.pkl')90 91# Load data92with open(labels_path, 'rb') as f:93    labels = pickle.load(f)94with open(predictions_path, 'rb') as f:95    predictions = pickle.load(f)96 97# Use cleanlab to analyze object detection data quality98summary = object_detection_health_summary(labels, predictions)99print(summary)100```101 102## Tutorial103 104For a complete tutorial using this dataset, see:105[Object Detection Tutorial](https://docs.cleanlab.ai/stable/tutorials/object_detection.html)106 107## Licensing Information108 109MIT License110 111## Citation112 113If you use this dataset in your research, please cite the cleanlab library:114 115```bibtex116@software{cleanlab,117  author = {Northcutt, Curtis G. and Athalye, Anish and Mueller, Jonas},118  title = {cleanlab},119  year = {2021},120  url = {https://github.com/cleanlab/cleanlab},121}122```123 124## Contact125 126- **Maintainers**: Cleanlab Team127- **Repository**: https://github.com/cleanlab/cleanlab128- **Documentation**: https://docs.cleanlab.ai129- **Issues**: https://github.com/cleanlab/cleanlab/issues130