snapshot...

modules/deeplearning/srcnn_l1b_l2.py

@@ -268,7 +268,7 @@ class SRCNN:
         # --------
         #tmp = input_data[:, label_idx, slc_y_2, slc_x_2]
         tmp = input_data[:, label_idx, :, :]
-        tmp = smooth_2d(tmp, sigma=1.5)
+        tmp = smooth_2d(tmp, sigma=1.0)
         tmp = tmp[:, slc_y_2, slc_x_2]
         if label_param != 'cloud_probability':
             tmp = normalize(tmp, label_param, mean_std_dct)
@@ -290,7 +290,7 @@ class SRCNN:
         # -----------------------------------------------------
         #label = input_data[:, label_idx, y_128, x_128]
         label = input_data[:, label_idx, :, :]
-        label = smooth_2d(label, sigma=1.5)
+        # label = smooth_2d(label, sigma=1.0)
         label = label[:, y_128, x_128]
 
         if label_param != 'cloud_probability':
@@ -762,7 +762,7 @@ def run_evaluate_static(in_file, out_file, ckpt_dir):
     # grd_b = normalize(grd_b, 'refl_0_65um_nom', mean_std_dct)
 
     grd_c = get_grid_values_all(h5f, label_param)
-    grd_c = gaussian_filter(grd_c, sigma=1.5)
+    # grd_c = gaussian_filter(grd_c, sigma=1.0)
     grd_c = grd_c[y_0:y_0+sub_y, x_0:x_0+sub_x]
     hr_grd_c = grd_c.copy()
     hr_grd_c = hr_grd_c[y_128, x_128]