ImageToTensor : Add view plug

include/GafferML/ImageToTensor.h

@@ -42,6 +42,7 @@
 #include "GafferImage/ImagePlug.h"
 
 #include "Gaffer/ComputeNode.h"
+#include "Gaffer/StringPlug.h"
 
 namespace GafferML
 {
@@ -59,6 +60,9 @@ class GAFFERML_API ImageToTensor : public Gaffer::ComputeNode
 		GafferImage::ImagePlug *imagePlug();
 		const GafferImage::ImagePlug *imagePlug() const;
 
+		Gaffer::StringPlug *viewPlug();
+		const Gaffer::StringPlug *viewPlug() const;
+
 		Gaffer::StringVectorDataPlug *channelsPlug();
 		const Gaffer::StringVectorDataPlug *channelsPlug() const;
 

python/GafferMLTest/ImageToTensorTest.py

@@ -36,6 +36,8 @@
 
 import unittest
 
+import imath
+
 import IECore
 
 import Gaffer
@@ -67,5 +69,40 @@ def testShufflingChannelsChangesHash( self ) :
 		tensor["channels"].setValue( IECore.StringVectorData( [ "B", "G", "R" ] ) )
 		self.assertNotEqual( tensor["tensor"].hash(), h1 )
 
+	def testView( self ) :
+
+		left = GafferImage.Constant()
+		left["color"].setValue( imath.Color4f( 1, 0, 0, 1 ) )
+		left["format"].setValue( GafferImage.Format( 1, 1 ) )
+
+		right = GafferImage.Constant()
+		right["color"].setValue( imath.Color4f( 0, 1, 0, 1 ) )
+		right["format"].setValue( GafferImage.Format( 1, 1 ) )
+
+		createViews = GafferImage.CreateViews()
+		createViews["views"].resize( 2 )
+		createViews["views"][0]["name"].setValue( "left" )
+		createViews["views"][0]["value"].setInput( left["out" ])
+		createViews["views"][1]["name"].setValue( "right" )
+		createViews["views"][1]["value"].setInput( right["out" ])
+
+		imageToTensor = GafferML.ImageToTensor()
+		imageToTensor["image"].setInput( createViews["out"] )
+
+		with self.assertRaisesRegex( Gaffer.ProcessException, "View does not exist" ) :
+			imageToTensor["tensor"].getValue()
+
+		imageToTensor["view"].setValue( "left" )
+		self.assertEqual(
+			imageToTensor["tensor"].getValue().asData(),
+			IECore.FloatVectorData( [ 1, 0, 0 ] )
+		)
+
+		imageToTensor["view"].setValue( "right" )
+		self.assertEqual(
+			imageToTensor["tensor"].getValue().asData(),
+			IECore.FloatVectorData( [ 0, 1, 0 ] )
+		)
+
 if __name__ == "__main__":
 	unittest.main()

python/GafferMLUI/ImageToTensorUI.py

@@ -0,0 +1,60 @@
+##########################################################################
+#
+#  Copyright (c) 2024, Cinesite VFX Ltd. All rights reserved.
+#
+#  Redistribution and use in source and binary forms, with or without
+#  modification, are permitted provided that the following conditions are
+#  met:
+#
+#      * Redistributions of source code must retain the above
+#        copyright notice, this list of conditions and the following
+#        disclaimer.
+#
+#      * Redistributions in binary form must reproduce the above
+#        copyright notice, this list of conditions and the following
+#        disclaimer in the documentation and/or other materials provided with
+#        the distribution.
+#
+#      * Neither the name of John Haddon nor the names of
+#        any other contributors to this software may be used to endorse or
+#        promote products derived from this software without specific prior
+#        written permission.
+#
+#  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS
+#  IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+#  THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+#  PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+#  CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+#  EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+#  PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+#  PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+#  LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+#  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+#  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+#
+##########################################################################
+
+import Gaffer
+import GafferML
+
+Gaffer.Metadata.registerNode(
+
+	GafferML.ImageToTensor,
+
+	# "description",
+	# """
+	# Converts Gaffer data to tensors for use with the Inference node.
+	# Potential data sources include PrimitiveVariableQuery nodes to fetch data
+	# from 3D scenes, or expressions to generate arbitrary input data.
+	# """,
+
+	plugs = {
+
+		"view" : [
+
+			"plugValueWidget:type", "GafferImageUI.ViewPlugValueWidget",
+
+		],
+
+	}
+)

python/GafferMLUI/__init__.py

@@ -35,6 +35,7 @@
 ##########################################################################
 
 from . import DataToTensorUI
+from . import ImageToTensorUI
 from . import TensorToImageUI
 from . import InferenceUI
 

src/GafferML/ImageToTensor.cpp

@@ -62,6 +62,7 @@ ImageToTensor::ImageToTensor( const std::string &name )
 {
 	storeIndexOfNextChild( g_firstPlugIndex );
 	addChild( new ImagePlug( "image", Plug::In ) );
+	addChild( new StringPlug( "view", Plug::In, "default" ) );
 	addChild( new StringVectorDataPlug( "channels", Plug::In, new StringVectorData( { "R", "G", "B" } ) ) );
 	addChild( new BoolPlug( "interleaveChannels" ) );
 	addChild( new TensorPlug( "tensor", Plug::Out ) );
@@ -81,44 +82,56 @@ const GafferImage::ImagePlug *ImageToTensor::imagePlug() const
 	return getChild<ImagePlug>( g_firstPlugIndex );
 }
 
+Gaffer::StringPlug *ImageToTensor::viewPlug()
+{
+	return getChild<StringPlug>( g_firstPlugIndex + 1 );
+}
+
+const Gaffer::StringPlug *ImageToTensor::viewPlug() const
+{
+	return getChild<StringPlug>( g_firstPlugIndex + 1 );
+}
+
 Gaffer::StringVectorDataPlug *ImageToTensor::channelsPlug()
 {
-	return getChild<StringVectorDataPlug>( g_firstPlugIndex + 1 );
+	return getChild<StringVectorDataPlug>( g_firstPlugIndex + 2 );
 }
 
 const Gaffer::StringVectorDataPlug *ImageToTensor::channelsPlug() const
 {
-	return getChild<StringVectorDataPlug>( g_firstPlugIndex + 1 );
+	return getChild<StringVectorDataPlug>( g_firstPlugIndex + 2 );
 }
 
 Gaffer::BoolPlug *ImageToTensor::interleaveChannelsPlug()
 {
-	return getChild<BoolPlug>( g_firstPlugIndex + 2 );
+	return getChild<BoolPlug>( g_firstPlugIndex + 3 );
 }
 
 const Gaffer::BoolPlug *ImageToTensor::interleaveChannelsPlug() const
 {
-	return getChild<BoolPlug>( g_firstPlugIndex + 2 );
+	return getChild<BoolPlug>( g_firstPlugIndex + 3 );
 }
 
 TensorPlug *ImageToTensor::tensorPlug()
 {
-	return getChild<TensorPlug>( g_firstPlugIndex + 3 );
+	return getChild<TensorPlug>( g_firstPlugIndex + 4 );
 }
 
 const TensorPlug *ImageToTensor::tensorPlug() const
 {
-	return getChild<TensorPlug>( g_firstPlugIndex + 3 );
+	return getChild<TensorPlug>( g_firstPlugIndex + 4 );
 }
 
 void ImageToTensor::affects( const Gaffer::Plug *input, AffectedPlugsContainer &outputs ) const
 {
 	ComputeNode::affects( input, outputs );
 
 	if(
+		input == imagePlug()->viewNamesPlug() ||
 		input == imagePlug()->dataWindowPlug() ||
 		input == imagePlug()->channelNamesPlug() ||
 		input == imagePlug()->channelDataPlug() ||
+		input == viewPlug() ||
 		input == channelsPlug() ||
 		input == interleaveChannelsPlug()
 	)
@@ -133,12 +146,16 @@ void ImageToTensor::hash( const Gaffer::ValuePlug *output, const Gaffer::Context
 	{
 		ComputeNode::hash( output, context, h );
 
-		const Box2i dataWindow = imagePlug()->dataWindow();
-		ConstStringVectorDataPtr inChannels = imagePlug()->channelNamesPlug()->getValue();
 		ConstStringVectorDataPtr channels = channelsPlug()->getValue();
-
 		interleaveChannelsPlug()->hash( h );
 
+		ImagePlug::ViewScope viewScope( context );
+		const std::string view = viewPlug()->getValue();
+		viewScope.setViewNameChecked( &view, imagePlug()->viewNames().get() );
+
+		const Box2i dataWindow = imagePlug()->dataWindow();
+		ConstStringVectorDataPtr inChannels = imagePlug()->channelNamesPlug()->getValue();
+
 		ImageAlgo::parallelGatherTiles(
 			imagePlug(),
 			channels->readable(),
@@ -162,7 +179,6 @@ void ImageToTensor::hash( const Gaffer::ValuePlug *output, const Gaffer::Context
 		);
 
 		h.append( dataWindow );
-
 	}
 	else
 	{
@@ -174,13 +190,16 @@ void ImageToTensor::compute( Gaffer::ValuePlug *output, const Gaffer::Context *c
 {
 	if( output == tensorPlug() )
 	{
-		const Box2i dataWindow = imagePlug()->dataWindow();
-		ConstStringVectorDataPtr inChannels = imagePlug()->channelNamesPlug()->getValue();
 		ConstStringVectorDataPtr channelsData = channelsPlug()->getValue();
 		const auto &channels = channelsData->readable();
-
 		const bool interleaveChannels = interleaveChannelsPlug()->getValue();
 
+		ImagePlug::ViewScope viewScope( context );
+		const std::string view = viewPlug()->getValue();
+		viewScope.setViewNameChecked( &view, imagePlug()->viewNames().get() );
+
+		const Box2i dataWindow = imagePlug()->dataWindow();
+		ConstStringVectorDataPtr inChannels = imagePlug()->channelNamesPlug()->getValue();
 		const size_t numPixels = dataWindow.size().x * dataWindow.size().y;
 
 		FloatVectorDataPtr bufferData = new FloatVectorData;