update link

Author: e-candeloro

README.md

@@ -44,7 +44,7 @@ The challenge comprises two primary tasks:
 
 # **Architecture and Inference Pipeline**
 
-Our approach is based on the state-of-the-art **CellViT-plus-plus** framework, which leverages a pre-trained foundational model backbone for **nuclei detection, segmentation, and classification** in whole slide images (WSIs). We enhance the system by fine-tuning a **multi-layer perceptron (MLP) classifier** to assign one of three classes to every detected nucleus: **monocytes, lymphocytes, and an additional "other" class**. The "other" class is generated semi-automatically using the **CellViT SAM-H model**, augmenting the training dataset for the **MONKEY challenge**.
+Our approach is based on the state-of-the-art **[CellViT-plus-plus](https://github.com/TIO-IKIM/CellViT-plus-plus)** framework, which leverages a pre-trained foundational model backbone for **nuclei detection, segmentation, and classification** in whole slide images (WSIs). We enhance the system by fine-tuning a **multi-layer perceptron (MLP) classifier** to assign one of three classes to every detected nucleus: **monocytes, lymphocytes, and an additional "other" class**. The "other" class is generated semi-automatically using the **CellViT SAM-H model**, augmenting the training dataset for the **MONKEY challenge**.
 
 ---
 

challenge_example_code

@@ -0,0 +1 @@
+Subproject commit 3a378b7d16663f21e061b27f7ad9ed61140c0cf2