diff --git a/mediapipe/tasks/python/metadata/metadata_writers/BUILD b/mediapipe/tasks/python/metadata/metadata_writers/BUILD
index 268dec49..e86254f2 100644
--- a/mediapipe/tasks/python/metadata/metadata_writers/BUILD
+++ b/mediapipe/tasks/python/metadata/metadata_writers/BUILD
@@ -57,6 +57,7 @@ py_library(
     deps = [
         ":metadata_info",
         ":metadata_writer",
+        ":writer_utils",
         "//mediapipe/tasks/metadata:image_segmenter_metadata_schema_py",
         "//mediapipe/tasks/metadata:metadata_schema_py",
         "//mediapipe/tasks/python/metadata",
diff --git a/mediapipe/tasks/python/metadata/metadata_writers/image_segmenter.py b/mediapipe/tasks/python/metadata/metadata_writers/image_segmenter.py
index 8e38731a..ee2dc092 100644
--- a/mediapipe/tasks/python/metadata/metadata_writers/image_segmenter.py
+++ b/mediapipe/tasks/python/metadata/metadata_writers/image_segmenter.py
@@ -22,6 +22,7 @@ from mediapipe.tasks.metadata import metadata_schema_py_generated as _metadata_f
 from mediapipe.tasks.python.metadata import metadata
 from mediapipe.tasks.python.metadata.metadata_writers import metadata_info
 from mediapipe.tasks.python.metadata.metadata_writers import metadata_writer
+from mediapipe.tasks.python.metadata.metadata_writers import writer_utils
 
 
 _MODEL_NAME = "ImageSegmenter"
@@ -148,10 +149,17 @@ class MetadataWriter(metadata_writer.MetadataWriterBase):
     writer = metadata_writer.MetadataWriter(model_buffer)
     writer.add_general_info(_MODEL_NAME, _MODEL_DESCRIPTION)
     writer.add_image_input(input_norm_mean, input_norm_std)
-    writer.add_segmentation_output(labels=labels)
     if activation is not None:
       option_md = ImageSegmenterOptionsMd(activation)
       writer.add_custom_metadata(option_md)
+    num_output_tensors = writer_utils.get_subgraph(model_buffer).OutputsLength()
+    if num_output_tensors == 2:
+      # For image segmenter model with 2 output tensors, the first one is
+      # quality score, and the second one is matting mask.
+      writer.add_feature_output(
+          "quality score", "The quality score of matting result."
+      )
+    writer.add_segmentation_output(labels=labels)
     return cls(writer)
 
   def populate(self) -> tuple[bytearray, str]: