Image Segmentation pipeline (#13828)

* Implement img seg pipeline

* Update src/transformers/pipelines/image_segmentation.py

Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Update src/transformers/pipelines/image_segmentation.py

Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

* Update output shape with individual masks

* Rm dev change

* Remove loops in test

Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>

docs/source/main_classes/pipelines.rst

@@ -29,6 +29,7 @@ There are two categories of pipeline abstractions to be aware about:
     - :class:`~transformers.FeatureExtractionPipeline`
     - :class:`~transformers.FillMaskPipeline`
     - :class:`~transformers.ImageClassificationPipeline`
+    - :class:`~transformers.ImageSegmentationPipeline`
     - :class:`~transformers.ObjectDetectionPipeline`
     - :class:`~transformers.QuestionAnsweringPipeline`
     - :class:`~transformers.SummarizationPipeline`
@@ -137,6 +138,13 @@ ImageClassificationPipeline
     :special-members: __call__
     :members:
 
+ImageSegmentationPipeline
+=======================================================================================================================
+
+.. autoclass:: transformers.ImageSegmentationPipeline
+    :special-members: __call__
+    :members:
+
 NerPipeline
 =======================================================================================================================
 

docs/source/model_doc/auto.rst

@@ -163,6 +163,13 @@ AutoModelForObjectDetection
     :members:
 
 
+AutoModelForImageSegmentation
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. autoclass:: transformers.AutoModelForImageSegmentation
+    :members:
+
+
 TFAutoModel
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~