Cross-dimensional Weighting for Aggregated Deep Convolutional Features

Benchmark Model Rank Results
image-retrieval-on-roxford-hardR – [O] –CroW#21mAP: 13.3
image-retrieval-on-roxford-mediumR – [O] –CroW#20mAP: 42.4
image-retrieval-on-rparis-hardR – [O] –CroW#14mAP: 47.2
image-retrieval-on-rparis-mediumR – [O] –CroW#14mAP: 70.4