diff --git a/.gitignore b/.gitignore
index 9953d9b..4460fd0 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,7 +4,6 @@ instant_test_output
 inference_test_output
 
 
-*.png
 *.json
 *.diff
 *.jpg
diff --git a/README.md b/README.md
index a3b4746..e56c060 100644
--- a/README.md
+++ b/README.md
@@ -2,7 +2,7 @@
 
 PAPS is a bottom approach for amodal panoptic segmentation, where the goal is to concurrently predict the pixel-wise semantic segmentation labels of visible regions of "stuff" classes (e.g., road, sky, and so on), and instance segmentation labels of both the visible and occluded regions of "thing" classes (e.g., car, truck, etc).
 
-![Overview of PAPS Architecture](/projects/paps/images/intro.png)
+![Overview of PAPS Architecture](/projects/PAPS/images/intro.png)
 
 This repository contains the **PyTorch implementation** of our RA-L'2022 paper [Perceiving the Invisible: Proposal-Free Amodal Panoptic Segmentation](https://arxiv.org/pdf/2205.14637.pdf). The repository builds on [Detectron2](https://github.com/facebookresearch/detectron2).
 
diff --git a/projects/PAPS/images/intro.png b/projects/PAPS/images/intro.png
deleted file mode 100644
index d9275cc..0000000
Binary files a/projects/PAPS/images/intro.png and /dev/null differ
diff --git a/projects/PAPS/images/overview.png b/projects/PAPS/images/overview.png
new file mode 100644
index 0000000..1662174
Binary files /dev/null and b/projects/PAPS/images/overview.png differ