diff --git a/src/symbols/symbol_utils.py b/src/symbols/symbol_utils.py
index a25d488..e23b01b 100644
--- a/src/symbols/symbol_utils.py
+++ b/src/symbols/symbol_utils.py
@@ -48,14 +48,19 @@ def get_fc1(last_conv, num_classes, fc_type):
       filters_in = num_classes
     else:
       body = last_conv
-    conv_6_sep = mx.sym.BatchNorm(data=body, fix_gamma=True, eps=2e-5, momentum=0.9, name='bn6f')  
-    spatial_norm=conv_6_sep*conv_6_sep
+    body = mx.sym.BatchNorm(data=body, fix_gamma=True, eps=2e-5, momentum=0.9, name='bn6f')  
+
+    spatial_norm=body*body
     spatial_norm=mx.sym.sum(data=spatial_norm, axis=1, keepdims=True)
     spatial_sqrt=mx.sym.sqrt(spatial_norm)
-    spatial_mean=mx.sym.mean(spatial_sqrt, axis=(1,2,3), keepdims=True)
-    #spatial_mean=mx.sym.mean(spatial_sqrt)
+    #spatial_mean=mx.sym.mean(spatial_sqrt, axis=(1,2,3), keepdims=True)
+    spatial_mean=mx.sym.mean(spatial_sqrt)
     spatial_div_inverse=mx.sym.broadcast_div(spatial_mean, spatial_sqrt)
-    body = mx.sym.broadcast_mul(body, spatial_div_inverse)
+
+    spatial_attention_inverse=mx.symbol.tile(spatial_div_inverse, reps=(1,filters_in,1,1))   
+    body=body*spatial_attention_inverse
+    #body = mx.sym.broadcast_mul(body, spatial_div_inverse)
+
     fc1 = mx.sym.Pooling(body, kernel=(7, 7), global_pool=True, pool_type='avg')
     if num_classes<filters_in:
       fc1 = mx.sym.BatchNorm(data=fc1, fix_gamma=True, eps=2e-5, momentum=0.9, name='bn6w')