Table 1 Number of images in Train, Valid, and Test sets per category.

From: Enhancing the YOLOv8 model for realtime object detection to ensure online platform safety

 

Datasets

Categories

Blood

Gun

Insulting Gesture

Knife

Alcohol

Cigarette

All

Images

Images

Images

Images

Images

Images

Images

Normal-cases

Train

388

701

328

1658

375

386

3836

Valid

83

149

69

354

79

82

816

Test

83

149

69

354

79

82

816

Subtotal

554

999

466

2366

533

550

5468

Hard-cases

Train

696

398

187

574

686

1078

3619

Valid

149

84

40

123

146

230

772

Test

149

84

40

123

146

230

772

Subtotal

994

566

267

820

978

1538

5163

Total

1548

1565

733

3186

1511

2088

10631