diff --git a/xlib/image/ImageProcessor.py b/xlib/image/ImageProcessor.py
index 6be6bae..a730625 100644
--- a/xlib/image/ImageProcessor.py
+++ b/xlib/image/ImageProcessor.py
@@ -839,19 +839,17 @@ class ImageProcessor:
         self._img = self._img.astype(np.uint8)
         return self
 
-    def to_dtype(self, dtype) -> 'ImageProcessor':
+    def to_dtype(self, dtype, from_tanh=False) -> 'ImageProcessor':
         if dtype == np.float32:
-            return self.to_ufloat32()
+            return self.to_ufloat32(from_tanh=from_tanh)
         elif dtype == np.uint8:
-            return self.to_uint8()
+            return self.to_uint8(from_tanh=from_tanh)
         else:
             raise ValueError('unsupported dtype')
 
-    def to_ufloat32(self, as_tanh=False) -> 'ImageProcessor':
+    def to_ufloat32(self, as_tanh=False, from_tanh=False) -> 'ImageProcessor':
         """
         Convert to uniform float32
-        if current image dtype uint8, then image will be divided by / 255.0
-        otherwise no operation
         """
         if self._img.dtype == np.uint8:
             self._img = self._img.astype(np.float32)
@@ -860,10 +858,14 @@ class ImageProcessor:
                 self._img -= 1.0
             else:
                 self._img /= 255.0
+        elif self._img.dtype in [np.float32, np.float64]:
+            if from_tanh:
+                self._img += 1.0
+                self._img /= 2.0
 
         return self
 
-    def to_uint8(self) -> 'ImageProcessor':
+    def to_uint8(self, from_tanh=False) -> 'ImageProcessor':
         """
         Convert to uint8
 
@@ -872,6 +874,10 @@ class ImageProcessor:
         img = self._img
 
         if img.dtype in [np.float32, np.float64]:
+            if from_tanh:
+                img += 1.0
+                img /= 2.0
+
             img *= 255.0
             np.clip(img, 0, 255, out=img)