fix

app.py

@@ -7,7 +7,7 @@ import sys
 import csv
 from monoscene.monoscene import MonoScene
 
-csv.field_size_limit(sys.maxsize)
+# csv.field_size_limit(sys.maxsize)
 torch.set_grad_enabled(False)
 
 # pipeline = pipeline(model="anhquancao/monoscene_kitti")
@@ -28,7 +28,6 @@ img_W, img_H = 1220, 370
 
 def predict(img):
     img = np.array(img, dtype=np.float32, copy=False) / 255.0
-    
 
     normalize_rgb = transforms.Compose(
             [
@@ -47,16 +46,17 @@ def predict(img):
 
     pred = model(batch).squeeze()
     # print(pred.shape)
-    pred = majority_pooling(pred, k_size=2)
-    fig = draw(pred, batch['fov_mask_2'])
+    # pred = majority_pooling(pred, k_size=2)
+    fig = draw(pred, batch['fov_mask_1'])
 
 
     return fig
    
+# The output is <b>downsampled by 2</b> to be able to be rendered in browsers. 
+
 description = """
 MonoScene Demo on SemanticKITTI Validation Set (Sequence 08), which uses the <b>camera parameters of Sequence 08</b>.
 Due to the <b>CPU-only</b> inference, it might take up to 20s to predict a scene. \n
-The output is <b>downsampled by 2</b> to be able to be rendered in browsers. 
 <b>Darker</b> colors represent the <b>scenery outside the Field of View</b>, i.e. not visible on the image.
 <center>
     <a href="https://cv-rits.github.io/MonoScene/">

helpers.py

@@ -188,7 +188,7 @@ def draw(
     fov_mask,
     # img_size,
     # f,
-    voxel_size=0.4,
+    voxel_size=0.2,
     # d=7,  # 7m - determine the size of the mesh representing the camera
 ):
 
@@ -254,7 +254,7 @@ def draw(
     # label = fov_voxels[:, 3].flatten()
     fig = go.Figure(data=[go.Scatter3d(x=x, y=y, z=z,mode='markers',
                     marker=dict(
-                            size=3,
+                            size=2,
                             color=pts_colors,                # set color to an array/list of desired values
                             # colorscale='Viridis',   # choose a colorscale
                             opacity=1.0,