Table 3.

Overview of the obtained results for our image retrieval experiments. In this table, we report the mAP@R, the R-Precision, and the Precision@1. The first 4 rows show the results on the ChestX-ray14 dataset for different image resolutions used for evaluation. The fifth row shows the outcomes on the CheXpert dataset. The last row indicates the results on the COVID-19 Image Data Collection. Bold text represents the overall highest performance metrics.

Dataset	Input dimensions	mAP@R	R-Precision	Precision@1
ChestX-ray14	1024 $\times$ 1024	0.9748	0.9763	0.9963
	800 $\times$ 800	0.9709	0.9726	0.9958
	512 $\times$ 512	0.9572	0.9601	0.9945
	224 $\times$ 224	0.7730	0.7979	0.9756
CheXpert	Original	0.8001	0.8148	0.9444
COVID-19	Original	0.8569	0.8707	0.8821