Learning and aggregating deep local descriptors for instance-level recognition

Open paper

Benchmark	Model	Rank	Results
image-retrieval-on-roxford-hard	HOW	#7	mAP: 56.9
image-retrieval-on-roxford-medium	HOW	#6	mAP: 79.4
image-retrieval-on-rparis-hard	HOW	#9	mAP: 62.4
image-retrieval-on-rparis-medium	HOW	#8	mAP: 81.6