app compatibile with streamlit 1.10

app.py

@@ -12,7 +12,7 @@ import torchvision
 from collections import OrderedDict
 from scipy import spatial
 import matplotlib.pyplot as plt
-
+from packaging import version
 
 from detectron2.engine import DefaultPredictor
 from detectron2.utils.visualizer import Visualizer
@@ -89,7 +89,6 @@ def load_model():
     cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = TH
     cfg.MODEL.DEVICE = "cpu"
 
-
     metadata = Metadata()
     metadata.set(
         evaluator_type="coco",
@@ -154,15 +153,22 @@ def explain(img, model):
     pred = cv2.resize(pred, (800, 800))
     pred = cv2.cvtColor(pred, cv2.COLOR_BGR2RGB)
     
-    tabs = st.tabs(["Result", "Detection"] + [f"Lesion #{i}" for i in range(0, len(instances))])
-    lesion_tabs = tabs[2:]
-
-    with tabs[0]:
-        st.header("Image processed")
-        st.success("Use the tabs on the right to see the detected lesions and detailed explanations for each lesion")
+    if version.parse(st.__version__) >= version.parse("1.11.0"):
+        tabs = st.tabs(["Result", "Detection"] + [f"Lesion #{i}" for i in range(0, len(instances))])
+        lesion_tabs = tabs[2:]
+        detection_tab = tabs[1]
+        with tabs[0]:
+            st.header("Image processed")
+            st.success("Use the tabs on the right to see the detected lesions and detailed explanations for each lesion")
+
+    else:
+        tabs = [st.container() for i in range(0, len(instances)+1)]
+        lesion_tabs = tabs[1:]
+        detection_tab = tabs[0]
+    
 
     state.write("Populating first tab...")
-    with tabs[1]:
+    with detection_tab:
         st.header("Detected lesions")
         st.image(pred)