From 5feb6d421e94c82ecb3112b933cddd00a92f7d72 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Caterina=20Fuster-Barcel=C3=B3?= <cafuster@pa.uc3m.es>
Date: Fri, 25 Aug 2023 10:23:24 +0200
Subject: [PATCH] Update README.md

---
 case2-finetuning/README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/case2-finetuning/README.md b/case2-finetuning/README.md
index ffe51ef..d20aac0 100644
--- a/case2-finetuning/README.md
+++ b/case2-finetuning/README.md
@@ -23,9 +23,9 @@ TODO: describe training in zerocost (@esgomezm)
 Integration with the multicut workflow enables cell segmentation based on the boundary predictions and allows to correct errors in the network prediction (that happen because of application to a different data modality) to be fixed by training an edge classifier.
 See the screenshots below for prediction and multicut results in ilastik.
 
-<img src="images/ilastik-prediction.png" alt="drawing" width="800"/>
-<img src="images/ilastik-mc.png" alt="drawing" width="800"/>
-<img src="images/ilastik-mc2.png" alt="drawing" width="800"/>
+<img src="https://github.com/bioimage-io/use-cases/blob/main/case2-finetuning/images/ilastik-prediction.png" alt="screenshot" width="800"/>
+<img src="https://github.com/bioimage-io/use-cases/blob/main/case2-finetuning/images/ilastik-mc.png" alt="screenshot" width="800"/>
+<img src="https://github.com/bioimage-io/use-cases/blob/main/case2-finetuning/images/ilastik-mc2.png" alt="screenshot" width="800"/>
 
 ## Retraining in ZeroCost