Hubness Precision Recall For Generative Image

project page: https://byronliang8.github.io/Hubness_Precision_Recall_page/

This is the code for paper and the hubness funcation CUDA version can used in HubnessGANSampling.

Requirements

• Linux and Windows are supported, but we recommend Linux for performance and compatibility reasons.
• 64-bit Python 3.7 and PyTorch 1.7.1. See https://pytorch.org/ for PyTorch install instructions.
• CUDA toolkit 11.0 or later. Use at least version 11.1 if running on RTX 3090.
• Python libraries: pip install click requests tqdm pyspng ninja imageio-ffmpeg==0.4.3 and pip install scikit-hubness.

Usage

This code is based on Github.

1. get the image features:

python getFeature.py datasetPath(real or fake images) --fname_precalc outputName.npz # The putput name should with .npz

2. compute precision and recall:

python improved_precision_recall_hubness.py real.npz fake.npz

Citation

@InProceedings{pmlr-v235-liang24f,
  title = 	 {Efficient Precision and Recall Metrics for Assessing Generative Models using Hubness-aware Sampling},
  author =       {Liang, Yuanbang and Wu, Jing and Lai, Yu-Kun and Qin, Yipeng},
  booktitle = 	 {Proceedings of the 41st International Conference on Machine Learning},
  pages = 	 {29682--29699},
  year = 	 {2024},
  volume = 	 {235},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {21--27 Jul},
  publisher =    {PMLR},
}