Simple image retrieval with mAP and PR curve.

You can use any image retrieval using deep learning. from github etc
It should work and the results should be reproducible.

The project is about
1- mAP (mean average precision).
https://towardsdatascience.com/breaking-down-mean-average-precision-map-ae462f623a52

2- Precision-Recall curve

Your code should calculate mAP for arbitrary number of query images.
and draw precision-recall curves from
https://github.com/lelan-li/SSAH/issues/5
or
https://github.com/swuxyj/DeepHash-pytorch/blob/master/utils/precision_recall_curve.py

Budget is $10-$20 and sorry, no negotiate.

Keras and TF.