Research Article
Automobile Component Recognition Based on Deep Learning Network with Coarse-Fine-Grained Feature Fusion
Table 6
Details of the accuracy improvement for PDLN on DATA_DEVICE by Algorithm
1 when
k = 1.2.
| | 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 |
| | Components | C0 | C1 | C2 | C3 | C4 | C5 | Average | | False ⟶ true | 52.94 | 54.55 | 0 | 25.00 | 16.13 | 39.47 | 33.51 | | True ⟶ false | 0 | 0 | 12.00 | 8.33 | 0 | 0 | 2.70 |
|
|
