Robust Keypoint Detection and Matching on Fisheye Images by Self-Supervised Learning

Table 3

Ablation study on different configurations of the proposed approach. The superscript denotes that the results are obtained at a threshold of 5 pixels. In the naive baseline, a hinge loss is adopted instead of the contrastive learning loss to learn descriptor correspondence. The best values are denoted in bold.

B	VT	CL	DC	RS	LE	HA	MS
✓				—		—	—
✓	✓			0.17	2.83	0.24	0.15
				0.35	2.73	0.37	0.39
			✓