diff --git a/src/TensorFlowNET.Keras/Layers/LayersApi.cs b/src/TensorFlowNET.Keras/Layers/LayersApi.cs
index 3cd46c28..dff9ff1b 100644
--- a/src/TensorFlowNET.Keras/Layers/LayersApi.cs
+++ b/src/TensorFlowNET.Keras/Layers/LayersApi.cs
@@ -66,7 +66,7 @@ namespace Tensorflow.Keras.Layers
                     Trainable = trainable,
                     Name = name
                 });
-
+#if false
         /// <summary>
         /// 2D convolution layer (e.g. spatial convolution over images).
         /// This layer creates a convolution kernel that is convolved with the layer input to produce a tensor of outputs.
@@ -119,6 +119,7 @@ namespace Tensorflow.Keras.Layers
                     ActivityRegularizer = activity_regularizer,
                     Activation = activation ?? keras.activations.Linear
                 });
+#endif
 
         /// <summary>
         /// 2D convolution layer (e.g. spatial convolution over images).
@@ -158,7 +159,7 @@ namespace Tensorflow.Keras.Layers
                 {
                     Rank = 2,
                     Filters = filters,
-                    KernelSize = kernel_size,
+                    KernelSize = (kernel_size == null) ? (5,5) : kernel_size,
                     Strides = strides == null ? (1, 1) : strides,
                     Padding = padding,
                     DataFormat = data_format,
diff --git a/test/TensorFlowNET.Keras.UnitTest/Layers/Layers.Convolution.Test.cs b/test/TensorFlowNET.Keras.UnitTest/Layers/Layers.Convolution.Test.cs
new file mode 100644
index 00000000..b24b214a
--- /dev/null
+++ b/test/TensorFlowNET.Keras.UnitTest/Layers/Layers.Convolution.Test.cs
@@ -0,0 +1,91 @@
+using Microsoft.VisualStudio.TestTools.UnitTesting;
+using NumSharp;
+using Tensorflow;
+using static Tensorflow.KerasApi;
+
+namespace TensorFlowNET.Keras.UnitTest
+{
+    [TestClass]
+    public class LayersConvolutionTest : EagerModeTestBase
+    {
+        [TestMethod]
+        public void BasicConv2D()
+        {
+            var filters = 8;
+            var conv = keras.layers.Conv2D(filters);
+
+            var x = np.arange(256.0f).reshape(1,8,8,4);
+            var y = conv.Apply(x);
+
+            Assert.AreEqual(4, y.shape.ndim);
+            Assert.AreEqual(x.shape[0], y.shape[0]);
+            Assert.AreEqual(x.shape[1] - 4, y.shape[1]);
+            Assert.AreEqual(x.shape[2] - 4, y.shape[2]);
+            Assert.AreEqual(filters, y.shape[3]);
+        }
+
+        [TestMethod]
+        public void BasicConv2D_ksize()
+        {
+            var filters = 8;
+            var conv = keras.layers.Conv2D(filters, kernel_size: 3);
+
+            var x = np.arange(256.0f).reshape(1, 8, 8, 4);
+            var y = conv.Apply(x);
+
+            Assert.AreEqual(4, y.shape.ndim);
+            Assert.AreEqual(x.shape[0], y.shape[0]);
+            Assert.AreEqual(x.shape[1] - 2, y.shape[1]);
+            Assert.AreEqual(x.shape[2] - 2, y.shape[2]);
+            Assert.AreEqual(filters, y.shape[3]);
+        }
+
+        [TestMethod]
+        public void BasicConv2D_ksize_same()
+        {
+            var filters = 8;
+            var conv = keras.layers.Conv2D(filters, kernel_size: 3, padding: "same");
+
+            var x = np.arange(256.0f).reshape(1, 8, 8, 4);
+            var y = conv.Apply(x);
+
+            Assert.AreEqual(4, y.shape.ndim);
+            Assert.AreEqual(x.shape[0], y.shape[0]);
+            Assert.AreEqual(x.shape[1], y.shape[1]);
+            Assert.AreEqual(x.shape[2], y.shape[2]);
+            Assert.AreEqual(filters, y.shape[3]);
+        }
+
+        [TestMethod]
+        public void BasicConv2D_ksize_strides()
+        {
+            var filters = 8;
+            var conv = keras.layers.Conv2D(filters, kernel_size: 3, strides: 2);
+
+            var x = np.arange(256.0f).reshape(1, 8, 8, 4);
+            var y = conv.Apply(x);
+
+            Assert.AreEqual(4, y.shape.ndim);
+            Assert.AreEqual(x.shape[0], y.shape[0]);
+            Assert.AreEqual(x.shape[1] - 5, y.shape[1]);
+            Assert.AreEqual(x.shape[2] - 5, y.shape[2]);
+            Assert.AreEqual(filters, y.shape[3]);
+        }
+
+        [TestMethod]
+        public void BasicConv2D_ksize_dilation()
+        {
+            var filters = 8;
+            var conv = keras.layers.Conv2D(filters, kernel_size: 3, dilation_rate: 2);
+
+            var x = np.arange(256.0f).reshape(1, 8, 8, 4);
+            var y = conv.Apply(x);
+
+            Assert.AreEqual(4, y.shape.ndim);
+            Assert.AreEqual(x.shape[0], y.shape[0]);
+            Assert.AreEqual(x.shape[1] - 4, y.shape[1]);
+            Assert.AreEqual(x.shape[2] - 4, y.shape[2]);
+            Assert.AreEqual(filters, y.shape[3]);
+        }
+    }
+}