Update app.py

app.py

@@ -46,39 +46,39 @@ def save_tensor_as_png(tensor: torch.Tensor, path: str, dtype: torch.dtype | Non
         dtype = tensor.dtype
     Image.fromarray(tensor.to(dtype).cpu().numpy()).save(path)
 
-# def colorize_points_with_turbo_all_dims(points, method='norm',cmap='turbo'):
-#     """
-#     Assigns colors to 3D points using the 'turbo' colormap based on a scalar computed from all 3 dimensions.
-
-#     Args:
-#         points (np.ndarray): (N, 3) array of 3D points.
-#         method (str): Method for reducing 3D point to scalar. Options: 'norm', 'pca'.
-
-#     Returns:
-#         np.ndarray: (N, 3) RGB colors in [0, 1].
-#     """
-#     assert points.shape[1] == 3, "Input must be of shape (N, 3)"
-
-#     if method == 'norm':
-#         scalar = np.linalg.norm(points, axis=1)
-#     elif method == 'pca':
-#         # Project onto first principal component
-#         mean = points.mean(axis=0)
-#         centered = points - mean
-#         u, s, vh = np.linalg.svd(centered, full_matrices=False)
-#         scalar = centered @ vh[0]  # Project onto first principal axis
-#     else:
-#         raise ValueError(f"Unknown method '{method}'")
-
-#     # Normalize scalar to [0, 1]
-#     scalar_min, scalar_max = scalar.min(), scalar.max()
-#     normalized = (scalar - scalar_min) / (scalar_max - scalar_min + 1e-8)
-
-#     # Apply turbo colormap
-#     cmap = plt.colormaps.get_cmap(cmap)
-#     colors = cmap(normalized)[:, :3]  # Drop alpha
-
-#     return colors
+def colorize_points_with_turbo_all_dims(points, method='norm',cmap='turbo'):
+    """
+    Assigns colors to 3D points using the 'turbo' colormap based on a scalar computed from all 3 dimensions.
+
+    Args:
+        points (np.ndarray): (N, 3) array of 3D points.
+        method (str): Method for reducing 3D point to scalar. Options: 'norm', 'pca'.
+
+    Returns:
+        np.ndarray: (N, 3) RGB colors in [0, 1].
+    """
+    assert points.shape[1] == 3, "Input must be of shape (N, 3)"
+
+    if method == 'norm':
+        scalar = np.linalg.norm(points, axis=1)
+    elif method == 'pca':
+        # Project onto first principal component
+        mean = points.mean(axis=0)
+        centered = points - mean
+        u, s, vh = np.linalg.svd(centered, full_matrices=False)
+        scalar = centered @ vh[0]  # Project onto first principal axis
+    else:
+        raise ValueError(f"Unknown method '{method}'")
+
+    # Normalize scalar to [0, 1]
+    scalar_min, scalar_max = scalar.min(), scalar.max()
+    normalized = (scalar - scalar_min) / (scalar_max - scalar_min + 1e-8)
+
+    # Apply turbo colormap
+    cmap = plt.colormaps.get_cmap(cmap)
+    colors = cmap(normalized)[:, :3]  # Drop alpha
+
+    return colors
 
 def prep_for_rayst3r(img,depth_dict,mask):
     H, W = img.shape[:2]