CPPE-5: Medical Personal Protective Equipment Dataset
We present a new challenging dataset, CPPE - 5 (Medical Personal Protective Equipment), with the goal to allow the study of subordinate categorization of medical personal protective equipments, which is not possible with other popular data sets that focus on broad-level categories (such as PASCAL VOC, ImageNet, Microsoft COCO, OpenImages, etc). To make it easy for models trained on this dataset to be used in practical scenarios in complex scenes, our dataset mainly contains images that show complex scenes with several objects in each scene in their natural context. The image collection for this dataset focuses on: obtaining as many non-iconic images as possible and making sure all the images are real-life images, unlike other existing datasets in this area. Our dataset includes 5 object categories (coveralls, face shields, gloves, masks, and goggles), and each image is annotated with a set of bounding boxes and positive labels. We present a detailed analysis of the dataset in comparison to other popular broad category datasets as well as datasets focusing on personal protective equipments, we also find that at present there exist no such publicly available datasets. Finally, we also analyze performance and compare model complexities on baseline and state-of-the-art models for bounding box results. Our code, data, and trained models are available at https://git.io/cppe5-dataset.
PDF AbstractCode
Tasks
Datasets
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Object Detection | CPPE-5 | TridentNet | box AP | 52.9 | # 1 | |
AP50 | 85.1 | # 5 | ||||
AP75 | 58.3 | # 2 | ||||
APS | 42.6 | # 4 | ||||
APM | 41.3 | # 5 | ||||
APL | 62.6 | # 1 | ||||
Object Detection | CPPE-5 | SSD | box AP | 29.50 | # 16 | |
AP50 | 57.0 | # 16 | ||||
AP75 | 24.9 | # 16 | ||||
APS | 32.1 | # 12 | ||||
APM | 23.1 | # 16 | ||||
APL | 34.6 | # 16 | ||||
Object Detection | CPPE-5 | YOLOv3 | box AP | 38.5 | # 15 | |
AP50 | 79.4 | # 9 | ||||
AP75 | 35.3 | # 15 | ||||
APS | 23.1 | # 16 | ||||
APM | 28.4 | # 15 | ||||
APL | 49.0 | # 14 | ||||
Object Detection | CPPE-5 | RepPoints | box AP | 43.0 | # 14 | |
AP50 | 75.9 | # 13 | ||||
AP75 | 40.1 | # 14 | ||||
APS | 27.3 | # 15 | ||||
APM | 36.7 | # 11 | ||||
APL | 48.0 | # 15 | ||||
Object Detection | CPPE-5 | Sparse RCNN | box AP | 44.0 | # 12 | |
AP50 | 69.6 | # 15 | ||||
AP75 | 44.6 | # 13 | ||||
APS | 30.0 | # 13 | ||||
APM | 30.6 | # 14 | ||||
APL | 54.7 | # 9 | ||||
Object Detection | CPPE-5 | Faster RCNN | box AP | 44.0 | # 12 | |
AP50 | 73.8 | # 14 | ||||
AP75 | 47.8 | # 11 | ||||
APS | 30.0 | # 13 | ||||
APM | 34.7 | # 13 | ||||
APL | 52.5 | # 12 | ||||
Object Detection | CPPE-5 | FCOS | box AP | 44.4 | # 11 | |
AP50 | 79.5 | # 8 | ||||
AP75 | 45.9 | # 12 | ||||
APS | 36.7 | # 8 | ||||
APM | 39.2 | # 9 | ||||
APL | 51.7 | # 13 | ||||
Object Detection | CPPE-5 | Grid RCNN | box AP | 47.5 | # 10 | |
AP50 | 77.9 | # 10 | ||||
AP75 | 50.6 | # 9 | ||||
APS | 43.4 | # 3 | ||||
APM | 37.2 | # 10 | ||||
APL | 54.4 | # 10 | ||||
Object Detection | CPPE-5 | Deformable DETR | box AP | 48.0 | # 9 | |
AP50 | 76.9 | # 11 | ||||
AP75 | 52.8 | # 7 | ||||
APS | 36.4 | # 9 | ||||
APM | 35.2 | # 12 | ||||
APL | 53.9 | # 11 | ||||
Object Detection | CPPE-5 | FSAF | box AP | 49.2 | # 8 | |
AP50 | 84.7 | # 6 | ||||
AP75 | 48.2 | # 10 | ||||
APS | 45.3 | # 2 | ||||
APM | 39.6 | # 8 | ||||
APL | 56.7 | # 8 | ||||
Object Detection | CPPE-5 | Localization Distillation | box AP | 50.9 | # 7 | |
AP50 | 76.5 | # 12 | ||||
AP75 | 58.8 | # 1 | ||||
APS | 45.8 | # 1 | ||||
APM | 43.0 | # 2 | ||||
APL | 59.4 | # 6 | ||||
Object Detection | CPPE-5 | VarifocalNet | box AP | 51.0 | # 6 | |
AP50 | 82.6 | # 7 | ||||
AP75 | 56.7 | # 3 | ||||
APS | 39.0 | # 5 | ||||
APM | 42.1 | # 3 | ||||
APL | 58.8 | # 7 | ||||
Object Detection | CPPE-5 | RegNet | box AP | 51.3 | # 5 | |
AP50 | 85.3 | # 4 | ||||
AP75 | 51.8 | # 8 | ||||
APS | 35.7 | # 11 | ||||
APM | 41.1 | # 6 | ||||
APL | 60.5 | # 5 | ||||
Object Detection | CPPE-5 | Deformable Convolutional Network | box AP | 51.6 | # 4 | |
AP50 | 87.1 | # 2 | ||||
AP75 | 55.9 | # 4 | ||||
APS | 36.3 | # 10 | ||||
APM | 41.4 | # 4 | ||||
APL | 61.3 | # 2 | ||||
Object Detection | CPPE-5 | Double Heads | box AP | 52.0 | # 3 | |
AP50 | 87.3 | # 1 | ||||
AP75 | 55.2 | # 5 | ||||
APS | 38.6 | # 7 | ||||
APM | 41.0 | # 7 | ||||
APL | 60.8 | # 4 | ||||
Object Detection | CPPE-5 | Empirical Attention | box AP | 52.5 | # 2 | |
AP50 | 86.5 | # 3 | ||||
AP75 | 54.1 | # 6 | ||||
APS | 38.7 | # 6 | ||||
APM | 43.4 | # 1 | ||||
APL | 61.0 | # 3 |