Merge branch 'test-facenet' into 'master'

Test Facenet

See merge request !84

bob/bio/face/test/test_embeddings.py

@@ -14,7 +14,7 @@ def test_idiap_inceptionv2_msceleb():
 
     reference = bob.io.base.load(
         pkg_resources.resource_filename(
-            "bob.bio.face.test", "data/inception_resnet_v2_rgb.hdf5"
+            "bob.bio.face.test", "data/inception_resnet_v2_msceleb_rgb.hdf5"
         )
     )
     np.random.seed(10)
@@ -61,11 +61,18 @@ def test_idiap_inceptionv2_msceleb_memory_demanding():
 
 @is_library_available("tensorflow")
 def test_idiap_inceptionv2_casia():
-    from bob.bio.face.embeddings import InceptionResnetv2_Casia_CenterLoss_2018
+    from bob.bio.face.embeddings.tf2_inception_resnet import (
+        InceptionResnetv2_Casia_CenterLoss_2018,
+    )
 
+    reference = bob.io.base.load(
+        pkg_resources.resource_filename(
+            "bob.bio.face.test", "data/inception_resnet_v2_casia_rgb.hdf5"
+        )
+    )
     np.random.seed(10)
     transformer = InceptionResnetv2_Casia_CenterLoss_2018()
-    data = np.random.rand(3, 160, 160).astype("uint8")
+    data = (np.random.rand(3, 160, 160) * 255).astype("uint8")
     output = transformer.transform([data])[0]
     assert output.size == 128, output.shape
 
@@ -74,6 +81,7 @@ def test_idiap_inceptionv2_casia():
     transformer_sample = wrap(["sample"], transformer)
     output = [s.data for s in transformer_sample.transform([sample])][0]
 
+    np.testing.assert_allclose(output, reference.flatten(), rtol=1e-5, atol=1e-4)
     assert output.size == 128, output.shape
 
 
@@ -83,9 +91,14 @@ def test_idiap_inceptionv1_msceleb():
         InceptionResnetv1_MsCeleb_CenterLoss_2018,
     )
 
+    reference = bob.io.base.load(
+        pkg_resources.resource_filename(
+            "bob.bio.face.test", "data/inception_resnet_v1_msceleb_rgb.hdf5"
+        )
+    )
     np.random.seed(10)
     transformer = InceptionResnetv1_MsCeleb_CenterLoss_2018()
-    data = np.random.rand(3, 160, 160).astype("uint8")
+    data = (np.random.rand(3, 160, 160) * 255).astype("uint8")
     output = transformer.transform([data])[0]
     assert output.size == 128, output.shape
 
@@ -94,6 +107,7 @@ def test_idiap_inceptionv1_msceleb():
     transformer_sample = wrap(["sample"], transformer)
     output = [s.data for s in transformer_sample.transform([sample])][0]
 
+    np.testing.assert_allclose(output, reference.flatten(), rtol=1e-5, atol=1e-4)
     assert output.size == 128, output.shape
 
 
@@ -103,9 +117,14 @@ def test_idiap_inceptionv1_casia():
         InceptionResnetv1_Casia_CenterLoss_2018,
     )
 
+    reference = bob.io.base.load(
+        pkg_resources.resource_filename(
+            "bob.bio.face.test", "data/inception_resnet_v1_casia_rgb.hdf5"
+        )
+    )
     np.random.seed(10)
     transformer = InceptionResnetv1_Casia_CenterLoss_2018()
-    data = np.random.rand(3, 160, 160).astype("uint8")
+    data = (np.random.rand(3, 160, 160) * 255).astype("uint8")
     output = transformer.transform([data])[0]
     assert output.size == 128, output.shape
 
@@ -114,6 +133,7 @@ def test_idiap_inceptionv1_casia():
     transformer_sample = wrap(["sample"], transformer)
     output = [s.data for s in transformer_sample.transform([sample])][0]
 
+    np.testing.assert_allclose(output, reference.flatten(), rtol=1e-5, atol=1e-4)
     assert output.size == 128, output.shape
 
 
@@ -123,9 +143,14 @@ def test_facenet_sanderberg():
         FaceNetSanderberg_20170512_110547,
     )
 
+    reference = bob.io.base.load(
+        pkg_resources.resource_filename(
+            "bob.bio.face.test", "data/facenet_sandberg_20170512-110547.hdf5"
+        )
+    )
     np.random.seed(10)
     transformer = FaceNetSanderberg_20170512_110547()
-    data = np.random.rand(3, 160, 160).astype("uint8")
+    data = (np.random.rand(3, 160, 160) * 255).astype("uint8")
     output = transformer.transform([data])[0]
     assert output.size == 128, output.shape
 
@@ -133,6 +158,8 @@ def test_facenet_sanderberg():
     sample = Sample(data)
     transformer_sample = wrap(["sample"], transformer)
     output = [s.data for s in transformer_sample.transform([sample])][0]
+
+    np.testing.assert_allclose(output, reference.flatten(), rtol=1e-5, atol=1e-4)
     assert output.size == 128, output.shape