Easy_start added

Author: eclique

Easy_start.ipynb

@@ -1,2 +1,30 @@
 # RISE
-Detect model's attention
+This repository contains source code necessary to reproduce some of the main results in the paper:
+
+[Vitali Petsiuk](http://cs-people.bu.edu/vpetsiuk/), [Abir Das](http://cs-people.bu.edu/dasabir/), [Kate Saenko](http://ai.bu.edu/ksaenko.html) (BMVC, 2018) <br>
+[RISE: Randomized Input Sampling for Explanation of Black-box Models](https://arxiv.org/abs/1806.07421)
+
+**If you use this software in an academic article, please consider citing:**
+
+    @inproceedings{Petsiuk2018rise,
+      title = {RISE: Randomized Input Sampling for Explanation of Black-box Models},
+      author = {Vitali Petsiuk and Abir Das and Kate Saenko},
+      booktitle = {Proceedings of the British Machine Vision Conference (BMVC)},
+      year = {2018}
+    }
+
+For more information regarding the paper, please visit http://cs-people.bu.edu/vpetsiuk/rise/.
+
+## Method overview
+To generate a saliency map for model's prediction, RISE queries black-box model on multiple randomly masked versions of input.
+After all the queries are done we average all the masks with respect to their scores to produce the final saliency map. The idea behind this is that whenever a mask preserves important parts of the image it gets higher score, and consequently has a higher weight in the sum.
+![](https://eclique.github.io/rep-imgs/RISE/rise-overview.png)
+
+## Repository contents
+* The whole idea is implemented in [Easy_start](Easy_start.ipynb) notebook, it's done in Keras but is really easy to modify for any framework, since the method itself is model-agnostic.
+* [Saliency](Saliency.ipynb) notebook demonstrates the usage of RISE class optimized for PyTorch.
+* [Evaluation](Evaluation.ipynb) notebook displays another contribution of the paper: *Causal metrics*.
+
+## Examples
+![](https://eclique.github.io/rep-imgs/RISE/example.png)
+![](https://eclique.github.io/rep-imgs/RISE/goldish.gif)

README.md

@@ -33,6 +33,7 @@ def generate_masks(self, N, s, p1, savepath='masks.npy'):
         self.masks = torch.from_numpy(self.masks).float()
         self.masks = self.masks.cuda()
         self.N = N
+        self.p1 = p1
 
     def load_masks(self, filepath):
         self.masks = np.load(filepath)
@@ -54,7 +55,7 @@ def forward(self, x):
         CL = p.size(1)
         sal = torch.matmul(p.data.transpose(0, 1), self.masks.view(N, H * W))
         sal = sal.view((CL, H, W))
-        sal /= N
+        sal = sal / N / self.p1
         return sal