Personal Protective Equipment Detection (PPE)
Dataset
Dataset Name | Training Images | Validation Images | Class Labels | License |
---|---|---|---|---|
Kaggle Construction Site Safety Dataset (v3) | 2,605 | 114 | 0: Hardhat 1: Mask 2: NO-Hardhat 3: NO-Mask 4: NO-Safety Vest 5: Person 6: Safety Cone 7: Safety Vest 8: machinery 9: vehicle |
CC BY 4.0 |
Model Zoo
DeGirum's PPE detection model zoo consists of a variety of models gathered and trained using ultralytics repository.
• The N2X models are located at N2X models' zoo.
• The OPENVINO models are located at OPENVINO models' zoo.
• The RKNN models are located at RKNN models' zoo.
Model Architecture | Input Size | Precision | Runtime | Device Type | mAP 50-95 | mAP 50 | FPS |
---|---|---|---|---|---|---|---|
yolov8n_relu6_ppe | 640x640 | INT8 | N2X | ORCA1 | 0.3619 | 0.6818 | 95.1 |
yolov8n_relu6_ppe | 640x640 | INT8 | RKNN | RK3588 | 0.3617 | 0.6771 | -- |
yolov8n_relu6_ppe | 640x640 | INT8 | OPENVINO | CPU | 0.3618 | 0.6782 | -- |
yolov8n_relu6_ppe | 640x640 | FP32 | OPENVINO | CPU | 0.3635 | 0.6789 | -- |
yolov8s_relu6_ppe | 640x640 | INT8 | N2X | ORCA1 | 0.5712 | 0.8302 | 41.6 |
yolov8s_relu6_ppe | 640x640 | INT8 | RKNN | RK3588 | 0.5735 | 0.8274 | -- |
yolov8s_relu6_ppe | 640x640 | INT8 | OPENVINO | CPU | 0.5725 | 0.8266 | -- |
yolov8s_relu6_ppe | 640x640 | FP32 | OPENVINO | CPU | 0.5732 | 0.8303 | -- |