fix: Update dependency fft-conv-pytorch to torch-fftconv

batchgeneratorsv2/transforms/nnunet/random_binary_operator.py

@@ -3,7 +3,7 @@
 
 import numpy as np
 import torch
-from fft_conv_pytorch import fft_conv
+from torch_fftconv import fft_conv2d, fft_conv3d
 
 from batchgeneratorsv2.helpers.scalar_type import RandomScalar, sample_scalar
 from batchgeneratorsv2.transforms.base.basic_transform import ImageOnlyTransform
@@ -30,11 +30,12 @@ def binary_dilation_torch(input_tensor, structure_element):
         raise ValueError("Input tensor must be 2D (X, Y) or 3D (X, Y, Z).")
 
     # Perform the convolution
-    # if num_dims == 2:  # 2D convolution
-    #     output = F.conv2d(input_tensor.unsqueeze(0).unsqueeze(0), structure_element, padding='same')
-    # elif num_dims == 3:  # 3D convolution
-    #     output = F.conv3d(input_tensor.unsqueeze(0).unsqueeze(0), structure_element, padding='same')
-    output = torch.round(fft_conv(input_tensor.unsqueeze(0).unsqueeze(0), structure_element, padding='same'), decimals=0)
+    input_tensor = input_tensor.unsqueeze(0).unsqueeze(0)
+    if num_dims == 2:  # 2D convolution
+        output = fft_conv2d(input_tensor, structure_element, padding='same')
+    elif num_dims == 3:  # 3D convolution
+        output = fft_conv3d(input_tensor, structure_element, padding='same')
+    output = torch.round(output, decimals=0)
 
     # Threshold to get binary output
     output = output > 0

batchgeneratorsv2/transforms/noise/gaussian_blur.py

@@ -8,7 +8,7 @@
 
 from batchgeneratorsv2.helpers.scalar_type import RandomScalar, sample_scalar
 from batchgeneratorsv2.transforms.base.basic_transform import ImageOnlyTransform
-from fft_conv_pytorch import fft_conv
+from torch_fftconv import fft_conv1d, fft_conv2d, fft_conv3d
 
 
 def blur_dimension(img: torch.Tensor, sigma: float, dim_to_blur: int, force_use_fft: bool = None, truncate: float = 6):
@@ -31,8 +31,9 @@ def blur_dimension(img: torch.Tensor, sigma: float, dim_to_blur: int, force_use_
 
     # Dynamically set up padding, convolution operation, and kernel shape based on the number of spatial dimensions
     conv_ops = {1: conv1d, 2: conv2d, 3: conv3d}
+    fft_conv_ops = {1: fft_conv1d, 2: fft_conv2d, 3: fft_conv3d}
     if force_use_fft is not None:
-        conv_op = conv_ops[spatial_dims] if not force_use_fft else fft_conv
+        conv_op = conv_ops[spatial_dims] if not force_use_fft else fft_conv_ops[spatial_dims]
     else:
         conv_op = conv_ops[spatial_dims]
 

pyproject.toml

@@ -32,7 +32,7 @@ keywords = [
 dependencies = [
     "torch>=2.0.0",
     "numpy",
-    "fft-conv-pytorch",
+    "torch-fftconv",
     "batchgenerators>=0.25"
 ]