diff --git a/qkeras/utils.py b/qkeras/utils.py
index 20d7c9f6..b311c63f 100644
--- a/qkeras/utils.py
+++ b/qkeras/utils.py
@@ -270,7 +270,7 @@ def model_quantize(model,
     # Dense becomes QDense
     # Activation converts activation functions
 
-    if layer["class_name"] == "Dense":
+    if layer["class_name"] in ["Dense", "QDense"]:
       layer["class_name"] = "QDense"
       # needs to add kernel/bias quantizers
       kernel_quantizer = get_config(
@@ -290,7 +290,7 @@ def model_quantize(model,
       else:
         quantize_activation(layer_config, activation_bits)
 
-    elif layer["class_name"] in ["Conv1D", "Conv2D"]:
+    elif layer["class_name"] in ["Conv1D", "Conv2D", "QConv1D", "QConv2D"]:
       q_name = "Q" + layer["class_name"]
       layer["class_name"] = q_name
       # needs to add kernel/bias quantizers
@@ -311,7 +311,7 @@ def model_quantize(model,
       else:
         quantize_activation(layer_config, activation_bits)
 
-    elif layer["class_name"] == "DepthwiseConv2D":
+    elif layer["class_name"] in ["DepthwiseConv2D", "QDepthwiseConv2D"]:
       layer["class_name"] = "QDepthwiseConv2D"
       # needs to add kernel/bias quantizers
       depthwise_quantizer = get_config(quantizer_config, layer,
@@ -330,7 +330,7 @@ def model_quantize(model,
       else:
         quantize_activation(layer_config, activation_bits)
 
-    elif layer["class_name"] == "Activation":
+    elif layer["class_name"] in ["Activation", "QActivation"]:
       quantizer = get_config(quantizer_config, layer, "QActivation")
       # this is to avoid softmax from quantizing in autoq
       if quantizer is None:
@@ -351,7 +351,7 @@ def model_quantize(model,
         else:
           quantize_activation(layer_config, activation_bits)
 
-    elif layer["class_name"] == "BatchNormalization":
+    elif layer["class_name"] in ["BatchNormalization", "QBatchNormalization"]:
       layer["class_name"] = "QBatchNormalization"
       # needs to add kernel/bias quantizers
       gamma_quantizer = get_config(