SHAP-Based Interpretable Object Detection Method for Satellite Imagery

This is the author implementation of SHAP-Based Interpretable Object Detection Method for Satellite Imagery. The implementation of the object detection model (YOLOv3) is based on Pytorch_YOLOv3. The framework of the proposed method can be applied to any differentiable object detection model.

Performance

Visualization

Please see the paper for details on the results of the evaluation, regularization, and data selection methods.

Installation

Requirements

• Python 3.6.3+
• Numpy
• OpenCV
• Matplotlib
• Pytorch 1.2+
• Cython
• Cuda (verified as operable: v10.2)
• Captum (verified as operable: v0.4.1)

optional:

• tensorboard
• tensorboardX
• CuDNN

Download the original YOLOv3 weights

download the pretrained file from the author's project page:

$ mkdir weights
$ cd weights/
$ bash ../requirements/download_weights.sh

Usage

Please see the test.ipynb

Paper

SHAP-based Methods for Interpretable Object Detection in Satellite Imagery

Hiroki Kawauchi, Takashi Fuse

[Paper] [Original Implementation]