Update app.py

app.py

@@ -566,6 +566,12 @@ with gr.Blocks(title = "Hello",
                         
                     send_btn = gr.Button("Run cell detection")
 
+            
+            sample_list = os.listdir("./gradio_examples/jpegs")
+            #sample_list = [ ("./gradio_examples/jpegs/"+sample, [((5, 5, 10, 10), 'nothing')]) for sample in sample_list]
+            
+            print(sample_list)
+            sample_list = [ "./gradio_examples/jpegs/"+sample for sample in sample_list]
                         
             #gr.Examples(sample_list, fn = update_with_example, inputs=input_image, outputs =  [input_image, up_btn, output_image], examples_per_page=50, label = "Click on an example to try it")
             example_image = gr.Image(visible=False, type='filepath')
@@ -588,11 +594,7 @@ with gr.Blocks(title = "Hello",
             down_btn = gr.DownloadButton("Download distance map (.tif)", visible=True)            
             down_btn2 = gr.DownloadButton("Download cell detections (.csv)", visible=True)  
 
-    sample_list = os.listdir("./gradio_examples/jpegs")
-    #sample_list = [ ("./gradio_examples/jpegs/"+sample, [((5, 5, 10, 10), 'nothing')]) for sample in sample_list]
-    
-    print(sample_list)
-    sample_list = [ "./gradio_examples/jpegs/"+sample for sample in sample_list]
+
     #sample_list = []
     #for j in range(23):
     #    sample_list.append("samples/img%0.2d.png"%j)
@@ -603,16 +605,26 @@ with gr.Blocks(title = "Hello",
     #down_btn.click(download_function, None, [down_btn, down_btn2])
     
     gr.HTML("""<h4 style="color:white;"> Notes:<br> 
-                    <li>you can load and process 3D, single-channel tifs (Z,Y,X).
-                    <li>withour GPU access, cell detection might take ~120 seconds. 
-                    <li>locally OrganoidTracker wil run much faste even on CPU. 
+                    <li>You can load and process 3D tifs in the following dimensions: (T),Z,(C),Y,X. We automatically pick the first timepoint.
+                    <li>Without GPU access, cell detection might take ~30 seconds. 
+                    <li>Locally OrganoidTracker wil run faster: ~2 seconds per frame on a dedicated GPU, ~10 seconds on a CPU. 
                     </h4>""")
 
     gr.HTML("""<h4 style="color:white;"> Caveats:<br> 
-                    <li>for this demo, an agressive background subtraction step is implemented before prediction, which we find benefits most usecases. For transperency, users have to preprocess the data themselves in OrganoidTracker 2.0. 
-                    <li>withour GPU access, cell detection might take ~120 seconds. 
-                    <li>locally OrganoidTracker wil run much faste even on CPU. 
+                    <li>For this demo, an agressive background subtraction step is implemented before prediction, which we find benefits most usecases. For transperency, users have to preprocess the data themselves in OrganoidTracker 2.0. 
+                    <li>Because of incompatibilities between TensorFlow and HuggingFace the models here are trained with the upcoming PyTorch version of OrganoidTracker (currently in beta). There might be performance differences when using the TensorFlow-versions presented in our paper.
                     </h4>""")
     
+    gr.HTML("""<h4 style="color:white;"> References:<br> 
+                    <li>The blastocyst sample data is from the BlastoSPIM dataset (Nunley et al., Development, 2024): 
+                    href="https://blastospim.flatironinstitute.org/html/index1.html" target="_blank">[website]</a>,
+                    href=https://journals.biologists.com/dev/article/151/21/dev202817/362603/Nuclear-instance-segmentation-and-tracking-for target="_blank">[paper]</a>
+                    
+                    <li>The blastocyst sample data is from the Cell Tracking Challenge (Murray et al., Nature Methods, 2008): 
+                    href="https://celltrackingchallenge.net/3d-datasets/" target="_blank">[website]</a>,
+                    href=https://www.nature.com/articles/nmeth.1228 target="_blank">[paper]</a>
+             </h4>""")
+    
+    
                     
 demo.queue().launch()