# Copyright 2021 The TensorFlow Authors. All Rights Reserved.
#
# 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 image_embedder."""
import enum
from absl.testing import parameterized
import numpy as np
import tensorflow as tf
from tensorflow_lite_support.python.task.core import base_options as base_options_module
from tensorflow_lite_support.python.task.processor.proto import bounding_box_pb2
from tensorflow_lite_support.python.task.processor.proto import embedding_options_pb2
from tensorflow_lite_support.python.task.processor.proto import embedding_pb2
from tensorflow_lite_support.python.task.vision import image_embedder
from tensorflow_lite_support.python.task.vision.core import tensor_image
from tensorflow_lite_support.python.test import test_util
_BaseOptions = base_options_module.BaseOptions
_ImageEmbedder = image_embedder.ImageEmbedder
_ImageEmbedderOptions = image_embedder.ImageEmbedderOptions
class ModelFileType(enum.Enum):
FILE_CONTENT = 1
FILE_NAME = 2
class ImageEmbedderTest(parameterized.TestCase, tf.test.TestCase):
def setUp(self):
super().setUp()
self.model_path = test_util.get_test_data_path(
"mobilenet_v3_small_100_224_embedder.tflite")
def test_create_from_file_succeeds_with_valid_model_path(self):
# Creates with default option and valid model file successfully.
embedder = _ImageEmbedder.create_from_file(self.model_path)
self.assertIsInstance(embedder, _ImageEmbedder)
def test_create_from_options_succeeds_with_valid_model_path(self):
# Creates with options containing model file successfully.
base_options = _BaseOptions(file_name=self.model_path)
options = _ImageEmbedderOptions(base_options=base_options)
embedder = _ImageEmbedder.create_from_options(options)
self.assertIsInstance(embedder, _ImageEmbedder)
def test_create_from_options_fails_with_invalid_model_path(self):
# Invalid empty model path.
with self.assertRaisesRegex(
ValueError,
r"ExternalFile must specify at least one of 'file_content', "
r"'file_name' or 'file_descriptor_meta'."):
base_options = _BaseOptions(file_name="")
options = _ImageEmbedderOptions(base_options=base_options)
_ImageEmbedder.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(file_content=f.read())
options = _ImageEmbedderOptions(base_options=base_options)
embedder = _ImageEmbedder.create_from_options(options)
self.assertIsInstance(embedder, _ImageEmbedder)
@parameterized.parameters(
(False, False, False, ModelFileType.FILE_NAME, 0.932738, -0.20580328),
(True, False, False, ModelFileType.FILE_NAME, 0.932738, -0.0135661615),
(True, True, False, ModelFileType.FILE_CONTENT, 0.929717, 254),
(False, False, True, ModelFileType.FILE_CONTENT, 0.999914, -0.16619979),
)
def test_embed(self, l2_normalize, quantize, with_bounding_box,
model_file_type, expected_similarity, expected_first_value):
# Creates embedder.
if model_file_type is ModelFileType.FILE_NAME:
base_options = _BaseOptions(file_name=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(file_content=model_content)
else:
# Should never happen
raise ValueError("model_file_type is invalid.")
embedding_options = embedding_options_pb2.EmbeddingOptions(
l2_normalize=l2_normalize, quantize=quantize)
options = _ImageEmbedderOptions(
base_options=base_options, embedding_options=embedding_options)
embedder = _ImageEmbedder.create_from_options(options)
# Loads images: one is a crop of the other.
image = tensor_image.TensorImage.create_from_file(
test_util.get_test_data_path("burger.jpg"))
cropped_image = tensor_image.TensorImage.create_from_file(
test_util.get_test_data_path("burger_crop.jpg"))
bounding_box = None
if with_bounding_box:
# Bounding box in "burger.jpg" corresponding to "burger_crop.jpg".
bounding_box = bounding_box_pb2.BoundingBox(
origin_x=0, origin_y=0, width=400, height=325)
# Extracts both embeddings.
image_result = embedder.embed(image, bounding_box)
crop_result = embedder.embed(cropped_image)
# Checks results sizes.
self.assertLen(image_result.embeddings, 1)
image_feature_vector = image_result.embeddings[0].feature_vector
self.assertLen(crop_result.embeddings, 1)
crop_feature_vector = crop_result.embeddings[0].feature_vector
self.assertLen(image_feature_vector.value, 1024)
self.assertLen(crop_feature_vector.value, 1024)
if quantize:
self.assertEqual(image_feature_vector.value.dtype, np.uint8)
else:
self.assertEqual(image_feature_vector.value.dtype, float)
# Check embedding value.
self.assertAlmostEqual(image_feature_vector.value[0], expected_first_value)
# Checks cosine similarity.
similarity = embedder.cosine_similarity(image_feature_vector,
crop_feature_vector)
self.assertAlmostEqual(similarity, expected_similarity, places=6)
def test_get_embedding_by_index(self):
base_options = _BaseOptions(file_name=self.model_path)
options = _ImageEmbedderOptions(base_options=base_options)
embedder = _ImageEmbedder.create_from_options(options)
# Builds test data.
feature_vector = embedding_pb2.FeatureVector(value=np.array([1.0, 0.0]))
embedding = embedding_pb2.Embedding(
output_index=0, feature_vector=feature_vector)
embedding_result = embedding_pb2.EmbeddingResult(embeddings=[embedding])
result0 = embedder.get_embedding_by_index(embedding_result, 0)
self.assertEqual(result0.output_index, 0)
self.assertEqual(result0.feature_vector.value[0], 1.0)
self.assertEqual(result0.feature_vector.value[1], 0.0)
with self.assertRaisesRegex(ValueError, r"Output index is out of bound\."):
embedder.get_embedding_by_index(embedding_result, 1)
def test_get_embedding_dimension(self):
base_options = _BaseOptions(file_name=self.model_path)
options = _ImageEmbedderOptions(base_options=base_options)
embedder = _ImageEmbedder.create_from_options(options)
self.assertEqual(embedder.get_embedding_dimension(0), 1024)
self.assertEqual(embedder.get_embedding_dimension(1), -1)
def test_number_of_output_layers(self):
base_options = _BaseOptions(file_name=self.model_path)
options = _ImageEmbedderOptions(base_options=base_options)
embedder = _ImageEmbedder.create_from_options(options)
self.assertEqual(embedder.number_of_output_layers, 1)
if __name__ == "__main__":
tf.test.main()
Codebase Browser
chromium