Classification of combustion state of sintering flame based on CNN-Transformer dual-stream network

Table 1. Comparison of learning effects of CNN-Transformer network models under different parameters
View table
View in Article
Table 1. Comparison of learning effects of CNN-Transformer network models under different parameters
Optimizers Learning rate Accuracy rate/%
Training set Validation set
SGD 0.01 99.56 95.07
0.001 99.85 96.20
0.000 1 99.37 94.36
Adam 0.01 98.67 85.69
0.001 99.23 92.35
0.000 1 98.26 78.96

Table 2. Comparison of results of different feature fusion methods
View table
View in Article
Table 2. Comparison of results of different feature fusion methods
Methods Accuracy/%
TFN 92.53
LMF 93.02
FCU 95.14
C-FCU 96.20

Table 3. Ablation experiment of CNN-Transformer algorithm
View table
View in Article
Table 3. Ablation experiment of CNN-Transformer algorithm
Models Accuracy/% Average accuracy/%
Normal flame Overburning flame Underburned flame
CNN 93.25 92.72 91.56 92.51
Transformer 95.00 93.45 93.55 94.00
CNN-Transformer 97.00 95.38 96.22 96.20

Table 4. Comparison of learning effects between CNN-Transformer dual-stream network model and other models

View table

View in Article

Table 4. Comparison of learning effects between CNN-Transformer dual-stream network model and other models

Models	Params/M	FLOPs/G	Accuracy/%	Training time/min	Speed/fs⁻¹
InceptionV3^[14]	23.21	6.02	90.50	85.72	20.79
ResNet18^[5]	11.65	3.82	91.00	67.95	22.53
ViT^[7]	55.50	77.91	90.49	157.65	12.24
MobileNet-V2^[15]	3.51	0.58	87.35	34.15	30.58
Conformer^[12]	23.53	5.23	94.35	73.85	23.86
CMT^[16]	25.10	4.02	95.00	76.29	20.56
Ours	12.74	3.54	96.20	61.25	25.78

Tools

Get Citation

Copy Citation Text

Xiuman LIANG, Jinming AN, Xiaohua CAO, Kai ZENG, Fubin WANG, Hefei LIU. Classification of combustion state of sintering flame based on CNN-Transformer dual-stream network[J]. Journal of Applied Optics, 2023, 44(5): 1030

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites