Jewelry Theft Detection
High-fidelity synthetic surveillance images capturing individuals interacting with jewelry displays and concealing merchandise. Diverse lighting and camera angles provide robust training data for detecting theft-related actions and suspicious bag usage in high-value retail jewelry store environments.
Sample Frames
20 annotated samples drawn from the train split. Toggle annotations to inspect bounding-box quality.
Class Distribution
3 annotation classes · 571 total images. Sorted by object count, descending.
Class Balance
Per-class counts, frequency, and average bounding-box area. Sort any column to surface the rarest or most prevalent classes for re-balancing.
| Class | Images↓ with class | Objects total | Per image average | Area % of frame |
|---|---|---|---|---|
| stealing | 211 | 211 | 1 | 8.30% |
| picking | 190 | 191 | 1.01 | 8.05% |
| person_with_bag | 171 | 179 | 1.05 | 7.21% |
Co-occurrence Matrix
How frequently pairs of classes appear in the same image. Diagonal cells show standalone image count for that class. Useful for spotting biased or correlated labels.
| stealing | picking | person_with_bag | |
|---|---|---|---|
| stealing | 211 | 1 | |
| picking | 1 | 190 | |
| person_with_bag | 171 |
Average Object Area
Each rectangle is one class, sized by the average area its bounding boxes occupy as a percentage of the frame. Surfaces tiny vs. dominant objects at a glance.
Spatial Distribution
Where annotations of each class tend to fall across the frame. Brighter regions indicate higher density — useful for detecting positional bias in your training data.
Model Performance
Validation metrics from a YOLOv8 detector trained on this dataset. Reference checkpoint: yolov8m.pt.
Validation curves over training
Dataset Metadata
| Annotation format | YOLO |
|---|---|
| Total images | 571 |
| Classes | 3 |
| Resolution | Mixed |
| Last updated | 2026-05 |
Interested in this dataset?
Get in touch for sample access, custom variants, or licensing terms — typical turnaround is two business days.