Merge pull request #4361 from kinaryml:face-aligner-python

PiperOrigin-RevId: 529165597
This commit is contained in:
Copybara-Service 2023-05-03 12:39:23 -07:00
commit e84e90e5b2
4 changed files with 384 additions and 0 deletions

View File

@ -185,3 +185,20 @@ py_test(
"@com_google_protobuf//:protobuf_python", "@com_google_protobuf//:protobuf_python",
], ],
) )
py_test(
name = "face_aligner_test",
srcs = ["face_aligner_test.py"],
data = [
"//mediapipe/tasks/testdata/vision:test_images",
"//mediapipe/tasks/testdata/vision:test_models",
],
deps = [
"//mediapipe/python:_framework_bindings",
"//mediapipe/tasks/python/components/containers:rect",
"//mediapipe/tasks/python/core:base_options",
"//mediapipe/tasks/python/test:test_utils",
"//mediapipe/tasks/python/vision:face_aligner",
"//mediapipe/tasks/python/vision/core:image_processing_options",
],
)

View File

@ -0,0 +1,190 @@
# Copyright 2023 The MediaPipe Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Tests for face aligner."""
import enum
import os
from absl.testing import absltest
from absl.testing import parameterized
from mediapipe.python._framework_bindings import image as image_module
from mediapipe.tasks.python.components.containers import rect
from mediapipe.tasks.python.core import base_options as base_options_module
from mediapipe.tasks.python.test import test_utils
from mediapipe.tasks.python.vision import face_aligner
from mediapipe.tasks.python.vision.core import image_processing_options as image_processing_options_module
_BaseOptions = base_options_module.BaseOptions
_Rect = rect.Rect
_Image = image_module.Image
_FaceAligner = face_aligner.FaceAligner
_FaceAlignerOptions = face_aligner.FaceAlignerOptions
_ImageProcessingOptions = image_processing_options_module.ImageProcessingOptions
_MODEL = 'face_landmarker_v2.task'
_LARGE_FACE_IMAGE = 'portrait.jpg'
_MODEL_IMAGE_SIZE = 256
_TEST_DATA_DIR = 'mediapipe/tasks/testdata/vision'
class ModelFileType(enum.Enum):
FILE_CONTENT = 1
FILE_NAME = 2
class FaceAlignerTest(parameterized.TestCase):
def setUp(self):
super().setUp()
self.test_image = _Image.create_from_file(
test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, _LARGE_FACE_IMAGE)
)
)
self.model_path = test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, _MODEL)
)
def test_create_from_file_succeeds_with_valid_model_path(self):
# Creates with default option and valid model file successfully.
with _FaceAligner.create_from_model_path(self.model_path) as aligner:
self.assertIsInstance(aligner, _FaceAligner)
def test_create_from_options_succeeds_with_valid_model_path(self):
# Creates with options containing model file successfully.
base_options = _BaseOptions(model_asset_path=self.model_path)
options = _FaceAlignerOptions(base_options=base_options)
with _FaceAligner.create_from_options(options) as aligner:
self.assertIsInstance(aligner, _FaceAligner)
def test_create_from_options_fails_with_invalid_model_path(self):
with self.assertRaisesRegex(
RuntimeError, 'Unable to open file at /path/to/invalid/model.tflite'
):
base_options = _BaseOptions(
model_asset_path='/path/to/invalid/model.tflite'
)
options = _FaceAlignerOptions(base_options=base_options)
_FaceAligner.create_from_options(options)
def test_create_from_options_succeeds_with_valid_model_content(self):
# Creates with options containing model content successfully.
with open(self.model_path, 'rb') as f:
base_options = _BaseOptions(model_asset_buffer=f.read())
options = _FaceAlignerOptions(base_options=base_options)
aligner = _FaceAligner.create_from_options(options)
self.assertIsInstance(aligner, _FaceAligner)
@parameterized.parameters(
(ModelFileType.FILE_NAME, _LARGE_FACE_IMAGE),
(ModelFileType.FILE_CONTENT, _LARGE_FACE_IMAGE),
)
def test_align(self, model_file_type, image_file_name):
# Load the test image.
self.test_image = _Image.create_from_file(
test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, image_file_name)
)
)
# Creates aligner.
if model_file_type is ModelFileType.FILE_NAME:
base_options = _BaseOptions(model_asset_path=self.model_path)
elif model_file_type is ModelFileType.FILE_CONTENT:
with open(self.model_path, 'rb') as f:
model_content = f.read()
base_options = _BaseOptions(model_asset_buffer=model_content)
else:
# Should never happen
raise ValueError('model_file_type is invalid.')
options = _FaceAlignerOptions(base_options=base_options)
aligner = _FaceAligner.create_from_options(options)
# Performs face alignment on the input.
alignd_image = aligner.align(self.test_image)
self.assertIsInstance(alignd_image, _Image)
# Closes the aligner explicitly when the aligner is not used in
# a context.
aligner.close()
@parameterized.parameters(
(ModelFileType.FILE_NAME, _LARGE_FACE_IMAGE),
(ModelFileType.FILE_CONTENT, _LARGE_FACE_IMAGE),
)
def test_align_in_context(self, model_file_type, image_file_name):
# Load the test image.
self.test_image = _Image.create_from_file(
test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, image_file_name)
)
)
# Creates aligner.
if model_file_type is ModelFileType.FILE_NAME:
base_options = _BaseOptions(model_asset_path=self.model_path)
elif model_file_type is ModelFileType.FILE_CONTENT:
with open(self.model_path, 'rb') as f:
model_content = f.read()
base_options = _BaseOptions(model_asset_buffer=model_content)
else:
# Should never happen
raise ValueError('model_file_type is invalid.')
options = _FaceAlignerOptions(base_options=base_options)
with _FaceAligner.create_from_options(options) as aligner:
# Performs face alignment on the input.
alignd_image = aligner.align(self.test_image)
self.assertIsInstance(alignd_image, _Image)
self.assertEqual(alignd_image.width, _MODEL_IMAGE_SIZE)
self.assertEqual(alignd_image.height, _MODEL_IMAGE_SIZE)
def test_align_succeeds_with_region_of_interest(self):
base_options = _BaseOptions(model_asset_path=self.model_path)
options = _FaceAlignerOptions(base_options=base_options)
with _FaceAligner.create_from_options(options) as aligner:
# Load the test image.
test_image = _Image.create_from_file(
test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, _LARGE_FACE_IMAGE)
)
)
# Region-of-interest around the face.
roi = _Rect(left=0.32, top=0.02, right=0.67, bottom=0.32)
image_processing_options = _ImageProcessingOptions(roi)
# Performs face alignment on the input.
alignd_image = aligner.align(test_image, image_processing_options)
self.assertIsInstance(alignd_image, _Image)
self.assertEqual(alignd_image.width, _MODEL_IMAGE_SIZE)
self.assertEqual(alignd_image.height, _MODEL_IMAGE_SIZE)
def test_align_succeeds_with_no_face_detected(self):
base_options = _BaseOptions(model_asset_path=self.model_path)
options = _FaceAlignerOptions(base_options=base_options)
with _FaceAligner.create_from_options(options) as aligner:
# Load the test image.
test_image = _Image.create_from_file(
test_utils.get_test_data_path(
os.path.join(_TEST_DATA_DIR, _LARGE_FACE_IMAGE)
)
)
# Region-of-interest that doesn't contain a human face.
roi = _Rect(left=0.1, top=0.1, right=0.2, bottom=0.2)
image_processing_options = _ImageProcessingOptions(roi)
# Performs face alignment on the input.
alignd_image = aligner.align(test_image, image_processing_options)
self.assertIsNone(alignd_image)
if __name__ == '__main__':
absltest.main()

View File

@ -264,3 +264,22 @@ py_library(
"//mediapipe/tasks/python/vision/core:vision_task_running_mode", "//mediapipe/tasks/python/vision/core:vision_task_running_mode",
], ],
) )
py_library(
name = "face_aligner",
srcs = [
"face_aligner.py",
],
deps = [
"//mediapipe/python:_framework_bindings",
"//mediapipe/python:packet_creator",
"//mediapipe/python:packet_getter",
"//mediapipe/tasks/cc/vision/face_stylizer/proto:face_stylizer_graph_options_py_pb2",
"//mediapipe/tasks/python/core:base_options",
"//mediapipe/tasks/python/core:optional_dependencies",
"//mediapipe/tasks/python/core:task_info",
"//mediapipe/tasks/python/vision/core:base_vision_task_api",
"//mediapipe/tasks/python/vision/core:image_processing_options",
"//mediapipe/tasks/python/vision/core:vision_task_running_mode",
],
)

View File

@ -0,0 +1,158 @@
# Copyright 2023 The MediaPipe Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""MediaPipe face aligner task."""
import dataclasses
from typing import Optional
from mediapipe.python import packet_creator
from mediapipe.python import packet_getter
from mediapipe.python._framework_bindings import image as image_module
from mediapipe.tasks.cc.vision.face_stylizer.proto import face_stylizer_graph_options_pb2
from mediapipe.tasks.python.core import base_options as base_options_module
from mediapipe.tasks.python.core import task_info as task_info_module
from mediapipe.tasks.python.core.optional_dependencies import doc_controls
from mediapipe.tasks.python.vision.core import base_vision_task_api
from mediapipe.tasks.python.vision.core import image_processing_options as image_processing_options_module
from mediapipe.tasks.python.vision.core import vision_task_running_mode as running_mode_module
_BaseOptions = base_options_module.BaseOptions
_FaceStylizerGraphOptionsProto = (
face_stylizer_graph_options_pb2.FaceStylizerGraphOptions
)
_RunningMode = running_mode_module.VisionTaskRunningMode
_ImageProcessingOptions = image_processing_options_module.ImageProcessingOptions
_TaskInfo = task_info_module.TaskInfo
_FACE_ALIGNMENT_IMAGE_NAME = 'face_alignment'
_FACE_ALIGNMENT_IMAGE_TAG = 'FACE_ALIGNMENT'
_NORM_RECT_STREAM_NAME = 'norm_rect_in'
_NORM_RECT_TAG = 'NORM_RECT'
_IMAGE_IN_STREAM_NAME = 'image_in'
_IMAGE_OUT_STREAM_NAME = 'image_out'
_IMAGE_TAG = 'IMAGE'
_TASK_GRAPH_NAME = 'mediapipe.tasks.vision.face_stylizer.FaceStylizerGraph'
@dataclasses.dataclass
class FaceAlignerOptions:
"""Options for the face aligner task.
Attributes:
base_options: Base options for the face aligner task.
"""
base_options: _BaseOptions
@doc_controls.do_not_generate_docs
def to_pb2(self) -> _FaceStylizerGraphOptionsProto:
"""Generates a FaceStylizerOptions protobuf object."""
base_options_proto = self.base_options.to_pb2()
base_options_proto.use_stream_mode = False
return _FaceStylizerGraphOptionsProto(base_options=base_options_proto)
class FaceAligner(base_vision_task_api.BaseVisionTaskApi):
"""Class that performs face alignment on images."""
@classmethod
def create_from_model_path(cls, model_path: str) -> 'FaceAligner':
"""Creates a `FaceAligner` object from a face landmarker task bundle and the default `FaceAlignerOptions`.
Note that the created `FaceAligner` instance is in image mode, for
aligning one face on a single image input.
Args:
model_path: Path to the face landmarker task bundle.
Returns:
`FaceAligner` object that's created from the model file and the default
`FaceAlignerOptions`.
Raises:
ValueError: If failed to create `FaceAligner` object from the provided
file such as invalid file path.
RuntimeError: If other types of error occurred.
"""
base_options = _BaseOptions(model_asset_path=model_path)
options = FaceAlignerOptions(base_options=base_options)
return cls.create_from_options(options)
@classmethod
def create_from_options(cls, options: FaceAlignerOptions) -> 'FaceAligner':
"""Creates the `FaceAligner` object from face aligner options.
Args:
options: Options for the face aligner task.
Returns:
`FaceAligner` object that's created from `options`.
Raises:
ValueError: If failed to create `FaceAligner` object from
`FaceAlignerOptions` such as missing the model.
RuntimeError: If other types of error occurred.
"""
task_info = _TaskInfo(
task_graph=_TASK_GRAPH_NAME,
input_streams=[
':'.join([_IMAGE_TAG, _IMAGE_IN_STREAM_NAME]),
':'.join([_NORM_RECT_TAG, _NORM_RECT_STREAM_NAME]),
],
output_streams=[
':'.join([_FACE_ALIGNMENT_IMAGE_TAG, _FACE_ALIGNMENT_IMAGE_NAME]),
':'.join([_IMAGE_TAG, _IMAGE_OUT_STREAM_NAME]),
],
task_options=options,
)
return cls(
task_info.generate_graph_config(enable_flow_limiting=False),
_RunningMode.IMAGE,
None,
)
def align(
self,
image: image_module.Image,
image_processing_options: Optional[_ImageProcessingOptions] = None,
) -> image_module.Image:
"""Performs face alignment on the provided MediaPipe Image.
Only use this method when the FaceAligner is created with the image
running mode.
Args:
image: MediaPipe Image.
image_processing_options: Options for image processing.
Returns:
The aligned face image. The aligned output image size is the same as the
model output size. None if no face is detected on the input image.
Raises:
ValueError: If any of the input arguments is invalid.
RuntimeError: If face alignment failed to run.
"""
normalized_rect = self.convert_to_normalized_rect(
image_processing_options, image
)
output_packets = self._process_image_data({
_IMAGE_IN_STREAM_NAME: packet_creator.create_image(image),
_NORM_RECT_STREAM_NAME: packet_creator.create_proto(
normalized_rect.to_pb2()
),
})
if output_packets[_FACE_ALIGNMENT_IMAGE_NAME].is_empty():
return None
return packet_getter.get_image(output_packets[_FACE_ALIGNMENT_IMAGE_NAME])