diff --git a/paddlespeech/t2s/frontend/zh_normalization/text_normlization.py b/paddlespeech/t2s/frontend/zh_normalization/text_normlization.py
index 9bdee638e0614977f6c0a7d40c613e357ff25d40..9794a70075bf79cbb978c83a4e9fa0bad3e76e1b 100644
--- a/paddlespeech/t2s/frontend/zh_normalization/text_normlization.py
+++ b/paddlespeech/t2s/frontend/zh_normalization/text_normlization.py
@@ -51,7 +51,7 @@ from .quantifier import replace_temperature
 
 class TextNormalizer():
     def __init__(self):
-        self.SENTENCE_SPLITOR = re.compile(r'([：、，；。？！,;?:!][”’]?)')
+        self.SENTENCE_SPLITOR = re.compile(r'([：、，；。？！,;?!][”’]?)')
 
     def _split(self, text: str, lang="zh") -> List[str]:
         """Split long text into sentences with sentence-splitting punctuations.