Update README.md

README.md

@@ -23,8 +23,7 @@ configs:
 
 [Computer Vision Lab, ETH Zurich](https://vision.ee.ethz.ch/)
 
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/682088962c40f64d03c4bff7/FhWwBRVvkFMtfdNQ-vnVT.png)
-
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/682088962c40f64d03c4bff7/P8z4Qtrw1-8R2IYQdnskJ.png)
 
 ## Introduction 
 We introduce the HardTracksDataset (HTD), a novel multi-object tracking (MOT) benchmark specifically designed to address two critical 
@@ -152,9 +151,20 @@ First install the HuggingFace CLI according to the official [HuggingFace documen
 and login with your HuggingFace account. Then, you can download the dataset using the following command:
 
 ```bash
-huggingface-cli download mscheidl/htd --repo-type dataset
+huggingface-cli download mscheidl/htd --repo-type dataset --local-dir htd
+```
+
+The video folders are provided as zip files. Before usage please unzip the files. You can use the following command to unzip all files in the `data` folder.
+Please note that the unzipping process can take a while (especially for _TAO.zip_)
+
+```bash
+cd htd
+for z in data/*.zip; do (unzip -o -q "$z" -d data && echo "Unzipped: $z") & done; wait; echo "✅ Done"
+mkdir -p data/zips        # create a folder for the zip files
+mv data/*.zip data/zips/  # move the zip files to the zips folder
 ```
 
+
 The dataset is organized in the following structure:
 
 ```