Metadata-Version: 2.1
Name: Xplique
Version: 0.1.0
Summary: Explanations toolbox for Tensorflow 2
Home-page: UNKNOWN
Author: Thomas FEL
Author-email: thomas.fel@brown.edu
License: MIT
Description: <div align="center">
            <img src="./docs/assets/typo.png" width="25%" alt="Xplique" align="center" />
        </div>
        <br>
        
        <div align="center">
            <a href="https://travis-ci.com/fel-thomas/xplique">
                <img alt="Build Status" src="https://travis-ci.com/fel-thomas/xplique.svg?token=R9xr216LTFpJW3LYYCaM&branch=master">
            </a>
        </div>
        <br>
        
        **Xplique** is a Python module dedicated to explainability. It provides several submodules to learn
        more about your tensorflow models (≥2.1). The three main submodules are _Attributions Methods_,
        _Explainability Metrics_ and _Feature Visualization_ tools.
        
        The _Attributions Method_ submodule implements various methods, with explanations, examples and 
        links to official papers.
        
        Soon, the _Explainability Metrics_ submodule will implement the current metrics related to 
        explainability. These evaluations used in conjunction with the attribution methods allow to measure
        the quality of the explanations.
        
        Soon, the _Feature Visualization_ submodule will allow to represent neurons, channels or layers
        by maximizing an input. 
        
        The package is released under [MIT license](https://choosealicense.com/licenses/mit).
        
        ![Example of Attributions Methods results](./docs/assets/samples.png)
        
        ## Contents
        
        - [Install](#installing) <br>
        - [Get started](#get-started) <br>
        - [Core features](#core-features) <br>
            - [Attributions Methods](#methods) <br>
            - [Concept based Methods](#concept-based-methods) <br>
            - [Metrics](#metrics) <br>
            - [Feature Visualization](#feature-visualization) <br>
        - [Notebooks](#notebooks) <br>
        
        ## Installing
        
        The library has been tested on Linux, MacOSX and Windows and relies on the following Python modules:
        
        * Tensorflow (>=2.1)
        * Numpy (>=1.18)
        
        You can install Xplique using pip with:
        
        ```bash
        pip install xplique
        ```
        
        ## Getting Started
        
        let's start with a simple example, by computing Grad-CAM for several images (or a complete dataset)
        on a trained model.
        
        ```python
        from xplique.attributions import GradCAM
        
        # load images, labels and model
        # ...
        
        method = GradCAM(model)
        explanations = method.explain(images, labels)
        ```
        
        ## Notebooks
        
        - [Using the attributions methods](https://gist.github.com/napolar/c02cef48ae7fc20e76d633f3f1588c63)
        <sub> [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/gist/napolar/c02cef48ae7fc20e76d633f3f1588c63/sample-generation.ipynb) </sub>
        
        ## Core features
        
        ### Attributions Methods
        
        * [x] Deconvolution          [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/deconvnet.md)               [📄<sup>arxiv</sup>](https://arxiv.org/abs/1311.2901)
        * [x] Grad-CAM               [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/grad_cam.md)                [📄<sup>arxiv</sup>](https://arxiv.org/abs/1610.02391)
        * [x] Grad-CAM++             [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/grad_cam_pp.md)             [📄<sup>arxiv</sup>](https://arxiv.org/abs/1710.11063)
        * [x] Gradient Input         [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/gradient_input.md)          [📄<sup>arxiv</sup>](https://arxiv.org/abs/1711.06104)
        * [x] Guided Backprop        [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/guided_backpropagation.md)  [📄<sup>arxiv</sup>](https://arxiv.org/abs/1412.6806)
        * [x] Integrated Gradients   [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/integrated_gradients.md)    [📄<sup>arxiv</sup>](https://arxiv.org/abs/1703.01365)
        * [x] Occlusion              [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/occlusion.md)               [📄<sup>arxiv</sup>](https://arxiv.org/abs/1311.2901)
        * [x] Rise                   [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/rise.md)                    [📄<sup>arxiv</sup>](https://arxiv.org/abs/1806.07421)   
        * [x] Saliency               [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/saliency.md)                [📄<sup>arxiv</sup>](https://arxiv.org/abs/1312.6034)
        * [x] SmoothGrad             [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/smoothgrad.md)              [📄<sup>arxiv</sup>](https://arxiv.org/abs/1706.03825)
        * [x] SquareGrad             [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/square_grad.md)             [📄<sup>arxiv</sup>](https://arxiv.org/abs/1806.10758)
        * [x] VarGrad                [ 📚<sup>Api</sup> ](https://napolar.github.io/xplique/api/vargrad.md)                 [📄<sup>arxiv</sup>](https://arxiv.org/abs/1810.03292)
        * [ ] Ablation-CAM  
        * [ ] Xray
        
        ### Concept-based Methods
        
        * [x] [ Concept Activation Vector ](./api/cav.md)[^12]
        * [x] [ Testing with Concept Activation Vector ](./api/tcav.md)[^12]
        * [ ] Robust TCAV 
        * [ ] Automatic Concept Extraction  
        
        ### Metrics
        
        * [ ] Aocp  
        * [ ] Fidelity correlation
        * [ ] Irof     
        * [ ] Pixel Flipping
        * [ ] Stability
        
        ### Feature Visualization
        
        * [ ] Vanilla
        
        
        
Platform: UNKNOWN
Classifier: Development Status :: 4 - Beta
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
Provides-Extra: tests
Provides-Extra: docs
