Implement BiRefNet and HAT Sharper models with background removal capabilities

- Added validation documentation for AI model tiers.
- Introduced `background.py` for background removal logic, including inference functions for alpha and foreground extraction.
- Created `onnx_birefnet.py` to handle BiRefNet model sessions and alpha inference.
- Developed tests for background models, ensuring proper dispatch and session reuse.
- Added tests for BiRefNet ONNX model to validate alpha output and error handling for invalid values.
- Implemented model export scripts for BiRefNet and HAT Sharper, ensuring proper weight compression and licensing.
- Created `deform_sampling.py` for deformable convolution operations used in BiRefNet.
- Added utility functions for compressing ONNX weights without altering graph arithmetic.
- Introduced a reference plane node for depth field adjustments in geometry processing.

tests/anyimage/operators/test_ai_input_failures.py

@@ -30,7 +30,7 @@ class AiInputFailuresTest(BlenderTestCase):
                     stack.enter_context(patch.object(self.runtime.runtime, "active_job", None))
                     submit = stack.enter_context(patch.object(self.runtime.runtime, "begin_job"))
                     if module == self.upscale:
-                        stack.enter_context(patch.object(module, "configured_upscale_model", return_value="SAFMN_REAL_X4"))
+                        stack.enter_context(patch.object(module, "configured_upscale_model", return_value="HAT_GAN_X4_SHARPER"))
                         stack.enter_context(patch.object(module, "configured_max_ai_input_size", return_value=2048))
                     self.assertEqual(operator.execute(context), {"CANCELLED"})
                     submit.assert_not_called()