Automobile Component Recognition Based on Deep Learning Network with Coarse-Fine-Grained Feature Fusion

Table 4

Ablation experiments of PDLN’s L and M.

Model	Training	Testing		Evaluation		Evaluation with Algorithm 1
	Dataset	Loss	Acc. (%)	Loss	Acc. (%)	Loss	Acc. (%)
PDLN_M	DATA_CON	0.0668	94.37	3.59709	55.14	1.976032	66.49
PDLN_L	DATA_CON	0.3581	92.81	5.9381	46.49	5.584	54.05
PDLN	DATA_CON	0.1945	95.11	1.7629	61.62	0.2841	92.43