Skip to main content

Table 6 Comparison of parameters and training time of all the models

From: FireXnet: an explainable AI-based tailored deep learning model for wildfire detection on resource-constrained devices

Models

Total parameters

Trainable parameters

Non-trainable parameters

Training time (HH:MM:SS)

Model size (MBs)

The proposed FireXnet model

1,774,308

37,988

1,736,320

01:13:10

6.73

DenseNet201

30,369,252

12,046,948

18,322,304

02:27:35

115.22

VGG16

17,930,980

3,215,972

14,715,008

01:49:37

67.95

MobileNetV2

10,291,172

8,032,868

2,258,304

02:10:35

39.12

InceptionResNetV2

59,256,324

4,919,588

54,336,736

01:49:26

224.92

InceptionV3

28,360,772

6,557,988

21,802,784

01:59:46

107.68