From 436e37609e31a5b78aee39b5c52e386f359284ea Mon Sep 17 00:00:00 2001
From: soltanianaref <101413285+soltanianaref@users.noreply.github.com>
Date: Tue, 5 Apr 2022 21:23:36 +0430
Subject: [PATCH] Update encoders.py

---
 steganogan/encoders.py | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/steganogan/encoders.py b/steganogan/encoders.py
index d1a93c8..84042fa 100644
--- a/steganogan/encoders.py
+++ b/steganogan/encoders.py
@@ -2,6 +2,14 @@
 
 import torch
 from torch import nn
+import torch.onnx
+from torchvision.ops.deform_conv import DeformConv2d
+
+input = torch.rand(4, 3, 10, 10)
+kh, kw = 3, 3
+weight = torch.rand(5, 3, kh, kw)
+offset = torch.rand(4, 2 * kh * kw, 8, 8)
+mask = torch.rand(4, kh * kw, 8, 8)
 
 
 class BasicEncoder(nn.Module):
@@ -16,11 +24,11 @@ class BasicEncoder(nn.Module):
     add_image = False
 
     def _conv2d(self, in_channels, out_channels):
-        return nn.Conv2d(
-            in_channels=in_channels,
-            out_channels=out_channels,
-            kernel_size=3,
-            padding=1
+        return DeformConv2d(
+            input,
+            offset=offset,
+            weight=weight,
+            mask=mask
         )
 
     def _build_models(self):