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

Table 5

The effect of different k for improving the recognition accuracy of PDLN on DATA_DEVICE.

False ⟶ true: recognition is wrong without Algorithm 1, and recognition is correct with Algorithm 1
True ⟶ false: recognition is correct without Algorithm 1, and recognition is wrong with Algorithm 1
K	1.0	1.1	1.2	1.3	1.4	1.5	1.6	1.7	2.0
False ⟶ true (%)	32.05	33.84	33.51	33.04	31.53	28.96	29.53	28.01	23.54
True ⟶ false (%)	17.02	12.27	2.70	8.19	5.72	4.62	4.24	3.83	3.31
Final Acc. (%)	71.35	76.76	92.43	78.92	78.46	78.38	78.92	77.84	73.51

The number that performed best in the multiple comparisons tested.