Unified Explanation Provider For CNNs
Project description
Explain LISA
It takes the following.
img: local path of img to be explained
class_names: the classes available as predictions for the given model
img_shape: shape of the image accepts by the neural network
model: the model to be explained get from tf.keras.models.load_model("your model path")
img1: local path background data point for produce explanations with SHAP
img2: local path background data point for produce explanations with SHAP
scale: for manual image scaling if scaling layer absent in the model to be explained
filter_radius: the pixel value of the radius of the High pass filter
Installation
pip install LISA_CNN_ExplainerV5
How to use it?
Open terminal and type python/python3 according to your OS.
import LISA_CNN_ExplainerV5 as e
e.ExplainLISA(img,class_names,img_shape,model,img1,img2,scale,filter_radius)
e.displayImages() \
License
© 2021 Sudil H.P Abeyagunasekera
This repository is licensed under the MIT license. See LICENSE for details.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for LISA_CNN_ExplainerV5-0.0.4.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6ca6fc1993b737ec04a77f457c599e6fc737f743a75706703e35d6601288f5e6 |
|
MD5 | a284f502dfeea6a0b0246949b26d8f6b |
|
BLAKE2b-256 | 2b7fc62ea18e21bf28a7a429bc38161fdbf2885bac461606f5b16c497aae5e5e |
Hashes for LISA_CNN_ExplainerV5-0.0.4-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2af79c22662851c3ff6116f54b5229170e8b961679c6ad58ec04bfc85d7bfad0 |
|
MD5 | 29e928e8302d55510a3bd534a76f6993 |
|
BLAKE2b-256 | 0015b79c73c442ebda7324a20051d8f0c798938132e4b7afc1362bc25e8a6181 |