| --- |
| license: cc-by-4.0 |
| task_categories: |
| - object-detection |
| language: |
| - en |
| pretty_name: Sensitive Words Detection Dataset |
| --- |
| |
| # Sensitive Words Detection Dataset |
|
|
| This dataset contains 101 images annotated for object detection of sensitive words. The annotations use YOLO text format, where each row is: |
|
|
| ```text |
| class_id x_center y_center width height |
| ``` |
|
|
| All coordinates are normalized to the image width and height. The only class is: |
|
|
| ```text |
| 0 sensitive_words |
| ``` |
|
|
| ## Files |
|
|
| - `images/`: image files. |
| - `labels/`: YOLO annotation files with the same basename as each image. |
| - `classes.txt`: class list. |
| - `data.yaml`: YOLO dataset configuration. |
| - `metadata.csv`: per-image metadata, including image size, label path, split, and box count. |
| - `notes.json`: Label Studio export metadata. |
|
|
| ## Dataset Statistics |
|
|
| - Images: 101 |
| - Label files: 101 |
| - Classes: 1 |
| - Bounding boxes: 830 |
| - Images with no annotated sensitive words: 8 |
|
|
| Empty label files indicate images with no annotated sensitive-word target. |
|
|
| ## Intended Use |
|
|
| This dataset is intended for research on visual detection of sensitive words in images, including object detection experiments and evaluation of OCR-adjacent content moderation systems. |
|
|
| ## Limitations |
|
|
| The dataset is small and should not be treated as representative of all languages, fonts, layouts, platforms, or sensitive-word policies. The class definition depends on the annotation guideline used during labeling, and downstream use should include additional validation before deployment. |
|
|
| ## Responsible AI Notes |
|
|
| Potential risks include false positives, false negatives, over-filtering of benign content, and policy bias if the dataset is used directly for automated moderation. This dataset should be used for research and evaluation rather than as the sole basis for high-stakes moderation decisions. |
|
|
| ## License |
|
|
| This dataset is released under the Creative Commons Attribution 4.0 International license (CC BY 4.0). |
|
|
