Update kwargs validation for `preprocess` with decorator (#32024)

* BLIP preprocess

* BIT preprocess

* BRIDGETOWER preprocess

* CHAMELEON preprocess

* CHINESE_CLIP preprocess

* CONVNEXT preprocess

* DEIT preprocess

* DONUT preprocess

* DPT preprocess

* FLAVA preprocess

* EFFICIENTNET preprocess

* FUYU preprocess

* GLPN preprocess

* IMAGEGPT preprocess

* INTRUCTBLIPVIDEO preprocess

* VIVIT preprocess

* ZOEDEPTH preprocess

* VITMATTE preprocess

* VIT preprocess

* VILT preprocess

* VIDEOMAE preprocess

* VIDEOLLAVA

* TVP processing

* TVP fixup

* SWIN2SR preprocess

* SIGLIP preprocess

* SAM preprocess

* RT-DETR preprocess

* PVT preprocess

* POOLFORMER preprocess

* PERCEIVER preprocess

* OWLVIT preprocess

* OWLV2 preprocess

* NOUGAT preprocess

* MOBILEVIT preprocess

* MOBILENETV2 preprocess

* MOBILENETV1 preprocess

* LEVIT preprocess

* LAYOUTLMV2 preprocess

* LAYOUTLMV3 preprocess

* Add test

* Update tests

tests/models/owlv2/test_image_processing_owlv2.py

@@ -47,6 +47,7 @@ class Owlv2ImageProcessingTester(unittest.TestCase):
         image_std=[0.26862954, 0.26130258, 0.27577711],
         do_convert_rgb=True,
     ):
+        super().__init__()
         self.parent = parent
         self.batch_size = batch_size
         self.num_channels = num_channels

tests/models/owlvit/test_image_processing_owlvit.py

@@ -44,6 +44,7 @@ class OwlViTImageProcessingTester(unittest.TestCase):
         image_std=[0.26862954, 0.26130258, 0.27577711],
         do_convert_rgb=True,
     ):
+        super().__init__()
         self.parent = parent
         self.batch_size = batch_size
         self.num_channels = num_channels

tests/models/poolformer/test_image_processing_poolformer.py

@@ -41,6 +41,7 @@ class PoolFormerImageProcessingTester(unittest.TestCase):
         image_mean=[0.5, 0.5, 0.5],
         image_std=[0.5, 0.5, 0.5],
     ):
+        super().__init__()
         size = size if size is not None else {"shortest_edge": 30}
         crop_size = crop_size if crop_size is not None else {"height": 30, "width": 30}
         self.parent = parent

tests/models/pvt/test_image_processing_pvt.py

@@ -41,6 +41,7 @@ class PvtImageProcessingTester(unittest.TestCase):
         image_mean=[0.485, 0.456, 0.406],
         image_std=[0.229, 0.224, 0.225],
     ):
+        super().__init__()
         size = size if size is not None else {"height": 18, "width": 18}
         self.parent = parent
         self.batch_size = batch_size

tests/models/rt_detr/test_image_processing_rt_detr.py

@@ -45,6 +45,7 @@ class RTDetrImageProcessingTester(unittest.TestCase):
         do_pad=False,
         return_tensors="pt",
     ):
+        super().__init__()
         self.parent = parent
         self.batch_size = batch_size
         self.num_channels = num_channels

tests/models/siglip/test_image_processing_siglip.py

@@ -43,6 +43,7 @@ class SiglipImageProcessingTester(unittest.TestCase):
         image_mean=[0.5, 0.5, 0.5],
         image_std=[0.5, 0.5, 0.5],
     ):
+        super().__init__()
         size = size if size is not None else {"height": 18, "width": 18}
         self.parent = parent
         self.batch_size = batch_size

tests/models/swin2sr/test_image_processing_swin2sr.py

@@ -48,6 +48,7 @@ class Swin2SRImageProcessingTester(unittest.TestCase):
         do_pad=True,
         pad_size=8,
     ):
+        super().__init__()
         self.parent = parent
         self.batch_size = batch_size
         self.num_channels = num_channels

tests/models/tvp/test_image_processing_tvp.py

@@ -58,6 +58,7 @@ class TvpImageProcessingTester(unittest.TestCase):
         num_channels=3,
         num_frames=2,
     ):
+        super().__init__()
         self.do_resize = do_resize
         self.size = size
         self.do_center_crop = do_center_crop

tests/models/video_llava/test_image_processing_video_llava.py

@@ -52,6 +52,7 @@ class VideoLlavaImageProcessingTester(unittest.TestCase):
         image_std=OPENAI_CLIP_STD,
         do_convert_rgb=True,
     ):
+        super().__init__()
         size = size if size is not None else {"shortest_edge": 20}
         crop_size = crop_size if crop_size is not None else {"height": 18, "width": 18}
         self.parent = parent

tests/models/videomae/test_image_processing_videomae.py

@@ -50,6 +50,7 @@ class VideoMAEImageProcessingTester(unittest.TestCase):
         image_std=[0.5, 0.5, 0.5],
         crop_size=None,
     ):
+        super().__init__()
         size = size if size is not None else {"shortest_edge": 18}
         crop_size = crop_size if crop_size is not None else {"height": 18, "width": 18}
 

tests/models/vilt/test_image_processing_vilt.py

@@ -46,6 +46,7 @@ class ViltImageProcessingTester(unittest.TestCase):
         image_mean=[0.5, 0.5, 0.5],
         image_std=[0.5, 0.5, 0.5],
     ):
+        super().__init__()
         size = size if size is not None else {"shortest_edge": 30}
         self.parent = parent
         self.batch_size = batch_size

tests/models/vit/test_image_processing_vit.py

@@ -44,6 +44,7 @@ class ViTImageProcessingTester(unittest.TestCase):
         image_mean=[0.5, 0.5, 0.5],
         image_std=[0.5, 0.5, 0.5],
     ):
+        super().__init__()
         size = size if size is not None else {"height": 18, "width": 18}
         self.parent = parent
         self.batch_size = batch_size

tests/models/vitmatte/test_image_processing_vitmatte.py

@@ -15,6 +15,7 @@
 
 
 import unittest
+import warnings
 
 import numpy as np
 
@@ -51,6 +52,7 @@ class VitMatteImageProcessingTester(unittest.TestCase):
         image_mean=[0.5, 0.5, 0.5],
         image_std=[0.5, 0.5, 0.5],
     ):
+        super().__init__()
         self.parent = parent
         self.batch_size = batch_size
         self.num_channels = num_channels
@@ -197,3 +199,20 @@ class VitMatteImageProcessingTest(ImageProcessingTestMixin, unittest.TestCase):
         image = np.random.randn(3, 249, 512)
         images = image_processing.pad_image(image)
         assert images.shape == (3, 256, 512)
+
+    def test_image_processor_preprocess_arguments(self):
+        # vitmatte require additional trimap input for image_processor
+        # that is why we override original common test
+
+        for image_processing_class in self.image_processor_list:
+            image_processor = image_processing_class(**self.image_processor_dict)
+            image = self.image_processor_tester.prepare_image_inputs()[0]
+            trimap = np.random.randint(0, 3, size=image.size[::-1])
+
+            with warnings.catch_warnings(record=True) as raised_warnings:
+                warnings.simplefilter("always")
+                image_processor(image, trimaps=trimap, extra_argument=True)
+
+            messages = " ".join([str(w.message) for w in raised_warnings])
+            self.assertGreaterEqual(len(raised_warnings), 1)
+            self.assertIn("extra_argument", messages)

tests/models/vivit/test_image_processing_vivit.py

@@ -50,6 +50,7 @@ class VivitImageProcessingTester(unittest.TestCase):
         image_std=[0.5, 0.5, 0.5],
         crop_size=None,
     ):
+        super().__init__()
         size = size if size is not None else {"shortest_edge": 18}
         crop_size = crop_size if crop_size is not None else {"height": 18, "width": 18}
 

tests/models/zoedepth/test_image_processing_zoedepth.py

@@ -46,6 +46,7 @@ class ZoeDepthImageProcessingTester(unittest.TestCase):
         image_std=[0.5, 0.5, 0.5],
         do_pad=False,
     ):
+        super().__init__()
         size = size if size is not None else {"height": 18, "width": 18}
         self.parent = parent
         self.batch_size = batch_size

tests/test_image_processing_common.py

@@ -19,6 +19,7 @@ import os
 import pathlib
 import tempfile
 import time
+import warnings
 
 import numpy as np
 import requests
@@ -425,8 +426,12 @@ class ImageProcessingTestMixin:
             )
 
     def test_image_processor_preprocess_arguments(self):
+        is_tested = False
+
         for image_processing_class in self.image_processor_list:
             image_processor = image_processing_class(**self.image_processor_dict)
+
+            # validation done by _valid_processor_keys attribute
             if hasattr(image_processor, "_valid_processor_keys") and hasattr(image_processor, "preprocess"):
                 preprocess_parameter_names = inspect.getfullargspec(image_processor.preprocess).args
                 preprocess_parameter_names.remove("self")
@@ -434,6 +439,28 @@ class ImageProcessingTestMixin:
                 valid_processor_keys = image_processor._valid_processor_keys
                 valid_processor_keys.sort()
                 self.assertEqual(preprocess_parameter_names, valid_processor_keys)
+                is_tested = True
+
+            # validation done by @filter_out_non_signature_kwargs decorator
+            if hasattr(image_processor.preprocess, "_filter_out_non_signature_kwargs"):
+                if hasattr(self.image_processor_tester, "prepare_image_inputs"):
+                    inputs = self.image_processor_tester.prepare_image_inputs()
+                elif hasattr(self.image_processor_tester, "prepare_video_inputs"):
+                    inputs = self.image_processor_tester.prepare_video_inputs()
+                else:
+                    self.skipTest(reason="No valid input preparation method found")
+
+                with warnings.catch_warnings(record=True) as raised_warnings:
+                    warnings.simplefilter("always")
+                    image_processor(inputs, extra_argument=True)
+
+                messages = " ".join([str(w.message) for w in raised_warnings])
+                self.assertGreaterEqual(len(raised_warnings), 1)
+                self.assertIn("extra_argument", messages)
+                is_tested = True
+
+        if not is_tested:
+            self.skipTest(reason="No validation found for `preprocess` method")
 
 
 class AnnotationFormatTestMixin: