Skip to content

Latest commit

 

History

History
30 lines (24 loc) · 766 Bytes

README.md

File metadata and controls

30 lines (24 loc) · 766 Bytes

NASA

Neighborhood-Adaptive Structure Augmented Metric Learning -- AAAI2022 oral

Pytorch implementation of NASA

How to use ?

1. Install requirements on your environment.

  • python=3.6.8
  • pytorch=1.4.0
  • numpy=1.16.0
  • tqdm=4.54.1
  • scipy
  • Pillow

2. Preparation.

mkdir ../MyDataset
  • Download data.

  • Extract the compressed file (tgz or zip) into ../MyDataset/, e.g., for Cars-196, put the files in the ../MyDataset/Cars196.

  • Download pre-train model.

3. Train.

python demo.py