Added Occlusion demo to README

- updated use of --relevance-norm

README.md

@@ -85,14 +85,15 @@ $ .venv/bin/python feed_forward.py \
     --parameters params/vgg16-397923af.pth \
     --model vgg16 \
     --composite epsilon_gamma_box \
+    --relevance-norm symmetric \
     --cmap coldnhot
 ```
 which computes the lrp heatmaps according to the `epsilon_gamma_box` rule and stores them in `results`, along with the respective input images.
 
 The resulting heatmaps may look like the following:
 ![beacon heatmaps](https://raw.githubusercontent.com/chr5tphr/zennit/master/share/img/beacon_vgg16_epsilon_gamma_box.png)
 
-Alternatively, heatmaps with absolute relevances for SmoothGrad may be computed by omitting `--composite` and supplying `--attributor`:
+Alternatively, heatmaps for SmoothGrad with absolute relevances may be computed by omitting `--composite` and supplying `--attributor`:
 ```shell
 $ .venv/bin/python feed_forward.py \
     data/lighthouses \
@@ -101,7 +102,21 @@ $ .venv/bin/python feed_forward.py \
     --parameters params/vgg16-397923af.pth \
     --model vgg16 \
     --attributor smoothgrad \
-    --absolute-relevance \
+    --relevance-norm absolute \
+    --cmap hot
+```
+For Integrated Gradients, `--attributor integrads` may be provided.
+
+Heatmaps for Occlusion Analysis with unaligned relevances may be computed by executing:
+```shell
+$ .venv/bin/python feed_forward.py \
+    data/lighthouses \
+    'results/vgg16_smoothgrad_{sample:02d}.png' \
+    --inputs 'results/vgg16_input_{sample:02d}.png' \
+    --parameters params/vgg16-397923af.pth \
+    --model vgg16 \
+    --attributor occlusion \
+    --relevance-norm unaligned \
     --cmap hot
 ```