IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[MO, nGraph, transforms., ref., python] Support DetectionOutput-8 (#8853)

Browse Source 
* Implement MO part for DetectionOutput-8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement nGraph part for DetectionOutput-8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement deduction of a number of classes method

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix build

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement nGraph tests to check shape inference for DetectionOutput-8 with deduction of num_classes

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement attributes tests for DetectionOutput-8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Remove DetectionOutput-1 from opset8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Add python bindings for DetectionOutput-8 and tests

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement downgrading and upgrading transformations for DetectionOutput

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Implement tests for downgrading and upgrading transformations for DetectionOutput

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Adopt existing reference for DetectionOutput-8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix DetectionOutput-8 reference

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Add tests for DetectionOutput-8 reference

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Finish resolving merge conflict

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix python bindings after re-structuring

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix BOM package list

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix multibox_detection_test

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix python bindings for DetectionOutput-8

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Insert blank line between summary and description for python binding

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Apply review comments: identation, for loop

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix formatting in python bindings

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix supported_attrs for DetectionOutput from opset1

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Add missed imports for python if_op

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Apply feedback and fix Win build

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Use the same test suite name for test mixture

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix transformation tests

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Recover BWDCMP_RTTI for legacy use and remove unused variables in tests

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Remove unused variables in type_prop tests

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Recover legacy referenceDetectionOutput constructor

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Fix python binding for DetectionOutput

Signed-off-by: Roman Kazantsev <roman.kazantsev@intel.com>
This commit is contained in:
Roman Kazantsev
2021-12-07 01:17:49 +03:00
committed by GitHub
parent b65a14d587
commit 42892767f5
44 changed files with 2189 additions and 844 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/bindings/python/src/compatibility/ngraph/opset8/__init__.py
View File

@@ -38,7 +38,7 @@ from ngraph.opset3.ops import cum_sum as cumsum
from ngraph.opset8.ops import deformable_convolution
from ngraph.opset1.ops import deformable_psroi_pooling
from ngraph.opset1.ops import depth_to_space
from ngraph.opset1.ops import detection_output
from ngraph.opset8.ops import detection_output
from ngraph.opset7.ops import dft
from ngraph.opset1.ops import divide
from ngraph.opset7.ops import einsum

228
src/bindings/python/src/compatibility/ngraph/opset8/ops.py
View File

@@ -3,41 +3,28 @@
"""Factory functions for all ngraph ops."""
from functools import partial
from typing import Callable, Iterable, List, Optional, Set, Union, Tuple
from typing import List, Optional, Tuple
import numpy as np
from ngraph.exceptions import UserInputError
from ngraph.impl import Node, Shape
from ngraph.impl.op import Constant, Parameter
from ngraph.impl import Node
from ngraph.opset_utils import _get_node_factory
from ngraph.utils.decorators import binary_op, nameable_op, unary_op
from ngraph.utils.decorators import nameable_op
from ngraph.utils.input_validation import (
assert_list_of_ints,
check_valid_attributes,
is_non_negative_value,
is_positive_value,
)
from ngraph.utils.node_factory import NodeFactory
from ngraph.utils.tensor_iterator_types import (
GraphBody,
TensorIteratorSliceInputDesc,
TensorIteratorMergedInputDesc,
TensorIteratorInvariantInputDesc,
TensorIteratorBodyOutputDesc,
TensorIteratorConcatOutputDesc,
)
from ngraph.utils.types import (
NodeInput,
NumericData,
NumericType,
ScalarData,
TensorShape,
as_node,
as_nodes,
get_dtype,
get_element_type,
get_element_type_str,
make_constant_node,
)
_get_node_factory_opset8 = partial(_get_node_factory, "opset8")
@@ -141,18 +128,18 @@ def adaptive_max_pool(
@nameable_op
def multiclass_nms(
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
iou_threshold: float = 0.0,
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
nms_eta: float = 1.0,
normalized: bool = True
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
iou_threshold: float = 0.0,
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
nms_eta: float = 1.0,
normalized: bool = True
) -> Node:
"""Return a node which performs MulticlassNms.
@@ -197,19 +184,19 @@ def multiclass_nms(
@nameable_op
def matrix_nms(
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
decay_function: str = "linear",
gaussian_sigma: float = 2.0,
post_threshold: float = 0.0,
normalized: bool = True
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
decay_function: str = "linear",
gaussian_sigma: float = 2.0,
post_threshold: float = 0.0,
normalized: bool = True
) -> Node:
"""Return a node which performs MatrixNms.
@@ -281,17 +268,17 @@ def gather(
@nameable_op
def max_pool(
data: NodeInput,
strides: List[int],
dilations: List[int],
pads_begin: List[int],
pads_end: List[int],
kernel_shape: TensorShape,
rounding_type: str = "floor",
auto_pad: Optional[str] = None,
index_element_type: Optional[str] = "i64",
axis: Optional[int] = 0,
name: Optional[str] = None,
data: NodeInput,
strides: List[int],
dilations: List[int],
pads_begin: List[int],
pads_end: List[int],
kernel_shape: TensorShape,
rounding_type: str = "floor",
auto_pad: Optional[str] = None,
index_element_type: Optional[str] = "i64",
axis: Optional[int] = 0,
name: Optional[str] = None,
) -> Node:
"""Perform max pooling operation and return both values and indices of the selected elements.
@@ -457,7 +444,7 @@ def gather_nd(
def prior_box(
layer_shape: Node, image_shape: NodeInput, attrs: dict, name: Optional[str] = None
layer_shape: Node, image_shape: NodeInput, attrs: dict, name: Optional[str] = None
) -> Node:
"""Generate prior boxes of specified sizes and aspect ratios across all dimensions.
@@ -655,3 +642,138 @@ def nv12_to_rgb(
inputs = as_nodes(arg, arg_uv)
return _get_node_factory_opset8().create("NV12toRGB", inputs)
@nameable_op
def detection_output(
box_logits: NodeInput,
class_preds: NodeInput,
proposals: NodeInput,
attrs: dict,
aux_class_preds: Optional[NodeInput] = None,
aux_box_preds: Optional[NodeInput] = None,
name: Optional[str] = None,
) -> Node:
"""Generate the detection output using information on location and confidence predictions.
@param box_logits: The 2D input tensor with box logits.
@param class_preds: The 2D input tensor with class predictions.
@param proposals: The 3D input tensor with proposals.
@param attrs: The dictionary containing key, value pairs for attributes.
@param aux_class_preds: The 2D input tensor with additional class predictions information.
@param aux_box_preds: The 2D input tensor with additional box predictions information.
@param name: Optional name for the output node.
@return Node representing DetectionOutput operation.
Available attributes are:
* background_label_id The background label id.
Range of values: integer value
Default value: 0
Required: no
* top_k Maximum number of results to be kept per batch after NMS step.
Range of values: integer value
Default value: -1
Required: no
* variance_encoded_in_target The flag that denotes if variance is encoded in target.
Range of values: {False, True}
Default value: False
Required: no
* keep_top_k Maximum number of bounding boxes per batch to be kept after NMS step.
Range of values: integer values
Default value: None
Required: yes
* code_type The type of coding method for bounding boxes.
Range of values: {'caffe.PriorBoxParameter.CENTER_SIZE',
'caffe.PriorBoxParameter.CORNER'}
Default value: 'caffe.PriorBoxParameter.CORNER'
Required: no
* share_location The flag that denotes if bounding boxes are shared among different
classes.
Range of values: {True, False}
Default value: True
Required: no
* nms_threshold The threshold to be used in the NMS stage.
Range of values: floating point value
Default value: None
Required: yes
* confidence_threshold Specifies the minimum confidence threshold for detection boxes to be
considered.
Range of values: floating point value
Default value: 0
Required: no
* clip_after_nms The flag that denotes whether to perform clip bounding boxes after
non-maximum suppression or not.
Range of values: {True, False}
Default value: False
Required: no
* clip_before_nms The flag that denotes whether to perform clip bounding boxes before
non-maximum suppression or not.
Range of values: {True, False}
Default value: False
Required: no
* decrease_label_id The flag that denotes how to perform NMS.
Range of values: False - perform NMS like in Caffe*.
True - perform NMS like in MxNet*.
Default value: False
Required: no
* normalized The flag that denotes whether input tensors with boxes are normalized.
Range of values: {True, False}
Default value: False
Required: no
* input_height The input image height.
Range of values: positive integer number
Default value: 1
Required: no
* input_width The input image width.
Range of values: positive integer number
Default value: 1
Required: no
* objectness_score The threshold to sort out confidence predictions.
Range of values: non-negative float number
Default value: 0
Required: no
Example of attribute dictionary:
@code{.py}
# just required ones
attrs = {
'keep_top_k': [1, 2, 3],
'nms_threshold': 0.645,
}
attrs = {
'keep_top_k': [1, 2, 3],
'nms_threshold': 0.645,
'normalized': True,
'clip_before_nms': True,
'input_height': [32],
'input_width': [32],
}
@endcode
Optional attributes which are absent from dictionary will be set with corresponding default.
"""
requirements = [
("background_label_id", False, np.integer, None),
("top_k", False, np.integer, None),
("variance_encoded_in_target", False, np.bool_, None),
("keep_top_k", True, np.integer, None),
("code_type", False, np.str_, None),
("share_location", False, np.bool_, None),
("nms_threshold", True, np.floating, None),
("confidence_threshold", False, np.floating, None),
("clip_after_nms", False, np.bool_, None),
("clip_before_nms", False, np.bool_, None),
("decrease_label_id", False, np.bool_, None),
("normalized", False, np.bool_, None),
("input_height", False, np.integer, is_positive_value),
("input_width", False, np.integer, is_positive_value),
("objectness_score", False, np.floating, is_non_negative_value),
]
check_valid_attributes("DetectionOutput", attrs, requirements)
inputs = [box_logits, class_preds, proposals]
if aux_class_preds is not None:
inputs.append(aux_class_preds)
if aux_box_preds is not None:
inputs.append(aux_box_preds)
inputs = as_nodes(*inputs)
return _get_node_factory_opset8().create("DetectionOutput", inputs, attrs)

2
src/bindings/python/src/openvino/runtime/opset8/__init__.py
View File

@@ -38,7 +38,7 @@ from openvino.runtime.opset3.ops import cum_sum as cumsum
from openvino.runtime.opset8.ops import deformable_convolution
from openvino.runtime.opset1.ops import deformable_psroi_pooling
from openvino.runtime.opset1.ops import depth_to_space
from openvino.runtime.opset1.ops import detection_output
from openvino.runtime.opset8.ops import detection_output
from openvino.runtime.opset7.ops import dft
from openvino.runtime.opset1.ops import divide
from openvino.runtime.opset7.ops import einsum

228
src/bindings/python/src/openvino/runtime/opset8/ops.py
View File

@@ -3,41 +3,28 @@
"""Factory functions for all openvino ops."""
from functools import partial
from typing import Callable, Iterable, List, Optional, Set, Union, Tuple
from typing import List, Optional, Tuple
import numpy as np
from openvino.runtime.exceptions import UserInputError
from openvino.runtime.impl import Node, Shape
from openvino.runtime.impl.op import Constant, Parameter
from openvino.runtime.impl import Node
from openvino.runtime.opset_utils import _get_node_factory
from openvino.runtime.utils.decorators import binary_op, nameable_op, unary_op
from openvino.runtime.utils.decorators import nameable_op
from openvino.runtime.utils.input_validation import (
assert_list_of_ints,
check_valid_attributes,
is_non_negative_value,
is_positive_value,
)
from openvino.runtime.utils.node_factory import NodeFactory
from openvino.runtime.utils.tensor_iterator_types import (
GraphBody,
TensorIteratorSliceInputDesc,
TensorIteratorMergedInputDesc,
TensorIteratorInvariantInputDesc,
TensorIteratorBodyOutputDesc,
TensorIteratorConcatOutputDesc,
)
from openvino.runtime.utils.types import (
NodeInput,
NumericData,
NumericType,
ScalarData,
TensorShape,
as_node,
as_nodes,
get_dtype,
get_element_type,
get_element_type_str,
make_constant_node,
)
_get_node_factory_opset8 = partial(_get_node_factory, "opset8")
@@ -141,18 +128,18 @@ def adaptive_max_pool(
@nameable_op
def multiclass_nms(
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
iou_threshold: float = 0.0,
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
nms_eta: float = 1.0,
normalized: bool = True
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
iou_threshold: float = 0.0,
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
nms_eta: float = 1.0,
normalized: bool = True
) -> Node:
"""Return a node which performs MulticlassNms.
@@ -197,19 +184,19 @@ def multiclass_nms(
@nameable_op
def matrix_nms(
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
decay_function: str = "linear",
gaussian_sigma: float = 2.0,
post_threshold: float = 0.0,
normalized: bool = True
boxes: NodeInput,
scores: NodeInput,
sort_result_type: str = "none",
sort_result_across_batch: bool = False,
output_type: str = "i64",
score_threshold: float = 0.0,
nms_top_k: int = -1,
keep_top_k: int = -1,
background_class: int = -1,
decay_function: str = "linear",
gaussian_sigma: float = 2.0,
post_threshold: float = 0.0,
normalized: bool = True
) -> Node:
"""Return a node which performs MatrixNms.
@@ -281,17 +268,17 @@ def gather(
@nameable_op
def max_pool(
data: NodeInput,
strides: List[int],
dilations: List[int],
pads_begin: List[int],
pads_end: List[int],
kernel_shape: TensorShape,
rounding_type: str = "floor",
auto_pad: Optional[str] = None,
index_element_type: Optional[str] = "i64",
axis: Optional[int] = 0,
name: Optional[str] = None,
data: NodeInput,
strides: List[int],
dilations: List[int],
pads_begin: List[int],
pads_end: List[int],
kernel_shape: TensorShape,
rounding_type: str = "floor",
auto_pad: Optional[str] = None,
index_element_type: Optional[str] = "i64",
axis: Optional[int] = 0,
name: Optional[str] = None,
) -> Node:
"""Perform max pooling operation and return both values and indices of the selected elements.
@@ -458,7 +445,7 @@ def gather_nd(
@nameable_op
def prior_box(
layer_shape: Node, image_shape: NodeInput, attrs: dict, name: Optional[str] = None
layer_shape: Node, image_shape: NodeInput, attrs: dict, name: Optional[str] = None
) -> Node:
"""Generate prior boxes of specified sizes and aspect ratios across all dimensions.
@@ -656,3 +643,138 @@ def nv12_to_rgb(
inputs = as_nodes(arg, arg_uv)
return _get_node_factory_opset8().create("NV12toRGB", inputs)
@nameable_op
def detection_output(
box_logits: NodeInput,
class_preds: NodeInput,
proposals: NodeInput,
attrs: dict,
aux_class_preds: Optional[NodeInput] = None,
aux_box_preds: Optional[NodeInput] = None,
name: Optional[str] = None,
) -> Node:
"""Generate the detection output using information on location and confidence predictions.
@param box_logits: The 2D input tensor with box logits.
@param class_preds: The 2D input tensor with class predictions.
@param proposals: The 3D input tensor with proposals.
@param attrs: The dictionary containing key, value pairs for attributes.
@param aux_class_preds: The 2D input tensor with additional class predictions information.
@param aux_box_preds: The 2D input tensor with additional box predictions information.
@param name: Optional name for the output node.
@return Node representing DetectionOutput operation.
Available attributes are:
* background_label_id The background label id.
Range of values: integer value
Default value: 0
Required: no
* top_k Maximum number of results to be kept per batch after NMS step.
Range of values: integer value
Default value: -1
Required: no
* variance_encoded_in_target The flag that denotes if variance is encoded in target.
Range of values: {False, True}
Default value: False
Required: no
* keep_top_k Maximum number of bounding boxes per batch to be kept after NMS step.
Range of values: integer values
Default value: None
Required: yes
* code_type The type of coding method for bounding boxes.
Range of values: {'caffe.PriorBoxParameter.CENTER_SIZE',
'caffe.PriorBoxParameter.CORNER'}
Default value: 'caffe.PriorBoxParameter.CORNER'
Required: no
* share_location The flag that denotes if bounding boxes are shared among different
classes.
Range of values: {True, False}
Default value: True
Required: no
* nms_threshold The threshold to be used in the NMS stage.
Range of values: floating point value
Default value: None
Required: yes
* confidence_threshold Specifies the minimum confidence threshold for detection boxes to be
considered.
Range of values: floating point value
Default value: 0
Required: no
* clip_after_nms The flag that denotes whether to perform clip bounding boxes after
non-maximum suppression or not.
Range of values: {True, False}
Default value: False
Required: no
* clip_before_nms The flag that denotes whether to perform clip bounding boxes before
non-maximum suppression or not.
Range of values: {True, False}
Default value: False
Required: no
* decrease_label_id The flag that denotes how to perform NMS.
Range of values: False - perform NMS like in Caffe*.
True - perform NMS like in MxNet*.
Default value: False
Required: no
* normalized The flag that denotes whether input tensors with boxes are normalized.
Range of values: {True, False}
Default value: False
Required: no
* input_height The input image height.
Range of values: positive integer number
Default value: 1
Required: no
* input_width The input image width.
Range of values: positive integer number
Default value: 1
Required: no
* objectness_score The threshold to sort out confidence predictions.
Range of values: non-negative float number
Default value: 0
Required: no
Example of attribute dictionary:
@code{.py}
# just required ones
attrs = {
'keep_top_k': [1, 2, 3],
'nms_threshold': 0.645,
}
attrs = {
'keep_top_k': [1, 2, 3],
'nms_threshold': 0.645,
'normalized': True,
'clip_before_nms': True,
'input_height': [32],
'input_width': [32],
}
@endcode
Optional attributes which are absent from dictionary will be set with corresponding default.
"""
requirements = [
("background_label_id", False, np.integer, None),
("top_k", False, np.integer, None),
("variance_encoded_in_target", False, np.bool_, None),
("keep_top_k", True, np.integer, None),
("code_type", False, np.str_, None),
("share_location", False, np.bool_, None),
("nms_threshold", True, np.floating, None),
("confidence_threshold", False, np.floating, None),
("clip_after_nms", False, np.bool_, None),
("clip_before_nms", False, np.bool_, None),
("decrease_label_id", False, np.bool_, None),
("normalized", False, np.bool_, None),
("input_height", False, np.integer, is_positive_value),
("input_width", False, np.integer, is_positive_value),
("objectness_score", False, np.floating, is_non_negative_value),
]
check_valid_attributes("DetectionOutput", attrs, requirements)
inputs = [box_logits, class_preds, proposals]
if aux_class_preds is not None:
inputs.append(aux_class_preds)
if aux_box_preds is not None:
inputs.append(aux_box_preds)
inputs = as_nodes(*inputs)
return _get_node_factory_opset8().create("DetectionOutput", inputs, attrs)

37
src/bindings/python/tests/test_ngraph/test_create_op.py
View File

@@ -110,7 +110,7 @@ def test_ctc_greedy_decoder(dtype):
(np.float64, np.int64, "i64", "i32", False, False),
(np.float64, np.int64, "i32", "i64", False, False),
(np.float64, np.int64, "i64", "i64", False, False)
],)
], )
def test_ctc_greedy_decoder_seq_len(fp_dtype, int_dtype, int_ci, int_sl, merge_repeated, blank_index):
input0_shape = [8, 20, 128]
input1_shape = [8]
@@ -1093,41 +1093,6 @@ def test_prior_box_clustered(int_dtype, fp_dtype):
assert list(node.get_output_shape(0)) == [2, 4332]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_detection_output(int_dtype, fp_dtype):
attributes = {
"num_classes": int_dtype(85),
"keep_top_k": np.array([64], dtype=int_dtype),
"nms_threshold": fp_dtype(0.645),
}
box_logits = ov.parameter([4, 8], fp_dtype, "box_logits")
class_preds = ov.parameter([4, 170], fp_dtype, "class_preds")
proposals = ov.parameter([4, 2, 10], fp_dtype, "proposals")
aux_class_preds = ov.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ov.parameter([4, 8], fp_dtype, "aux_box_preds")
node = ov.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_type_name() == "DetectionOutput"
assert node.get_output_size() == 1
assert list(node.get_output_shape(0)) == [1, 1, 256, 7]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[

111
src/bindings/python/tests/test_ngraph/test_detection_output.py Normal file
View File

@@ -0,0 +1,111 @@
# Copyright (C) 2018-2021 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
import numpy as np
import openvino.runtime.opset8 as ov
import pytest
np_types = [np.float32, np.int32]
integral_np_types = [
np.int8,
np.int16,
np.int32,
np.int64,
np.uint8,
np.uint16,
np.uint32,
np.uint64,
]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_detection_output(int_dtype, fp_dtype):
attributes = {
"keep_top_k": np.array([64], dtype=int_dtype),
"nms_threshold": fp_dtype(0.645),
}
box_logits = ov.parameter([4, 8], fp_dtype, "box_logits")
class_preds = ov.parameter([4, 170], fp_dtype, "class_preds")
proposals = ov.parameter([4, 2, 10], fp_dtype, "proposals")
aux_class_preds = ov.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ov.parameter([4, 8], fp_dtype, "aux_box_preds")
node = ov.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_type_name() == "DetectionOutput"
assert node.get_output_size() == 1
assert list(node.get_output_shape(0)) == [1, 1, 256, 7]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_dynamic_get_attribute_value(int_dtype, fp_dtype):
attributes = {
"background_label_id": int_dtype(13),
"top_k": int_dtype(16),
"variance_encoded_in_target": True,
"keep_top_k": np.array([64, 32, 16, 8], dtype=int_dtype),
"code_type": "caffe.PriorBoxParameter.CENTER_SIZE",
"share_location": False,
"nms_threshold": fp_dtype(0.645),
"confidence_threshold": fp_dtype(0.111),
"clip_after_nms": True,
"clip_before_nms": False,
"decrease_label_id": True,
"normalized": True,
"input_height": int_dtype(86),
"input_width": int_dtype(79),
"objectness_score": fp_dtype(0.77),
}
box_logits = ov.parameter([4, 680], fp_dtype, "box_logits")
class_preds = ov.parameter([4, 170], fp_dtype, "class_preds")
proposals = ov.parameter([4, 1, 8], fp_dtype, "proposals")
aux_class_preds = ov.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ov.parameter([4, 680], fp_dtype, "aux_box_preds")
node = ov.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_background_label_id() == int_dtype(13)
assert node.get_top_k() == int_dtype(16)
assert node.get_variance_encoded_in_target()
assert np.all(np.equal(node.get_keep_top_k(), np.array([64, 32, 16, 8], dtype=int_dtype)))
assert node.get_code_type() == "caffe.PriorBoxParameter.CENTER_SIZE"
assert not node.get_share_location()
assert np.isclose(node.get_nms_threshold(), fp_dtype(0.645))
assert np.isclose(node.get_confidence_threshold(), fp_dtype(0.111))
assert node.get_clip_after_nms()
assert not node.get_clip_before_nms()
assert node.get_decrease_label_id()
assert node.get_normalized()
assert node.get_input_height() == int_dtype(86)
assert node.get_input_width() == int_dtype(79)
assert np.isclose(node.get_objectness_score(), fp_dtype(0.77))

65
src/bindings/python/tests/test_ngraph/test_dyn_attributes.py
View File

@@ -2,9 +2,8 @@
# SPDX-License-Identifier: Apache-2.0
import numpy as np
import pytest
import openvino.runtime.opset8 as ov
import pytest
@pytest.fixture()
@@ -37,68 +36,6 @@ def test_dynamic_attributes_softmax():
assert node.get_axis() == 3
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_dynamic_get_attribute_value(int_dtype, fp_dtype):
attributes = {
"num_classes": int_dtype(85),
"background_label_id": int_dtype(13),
"top_k": int_dtype(16),
"variance_encoded_in_target": True,
"keep_top_k": np.array([64, 32, 16, 8], dtype=int_dtype),
"code_type": "caffe.PriorBoxParameter.CENTER_SIZE",
"share_location": False,
"nms_threshold": fp_dtype(0.645),
"confidence_threshold": fp_dtype(0.111),
"clip_after_nms": True,
"clip_before_nms": False,
"decrease_label_id": True,
"normalized": True,
"input_height": int_dtype(86),
"input_width": int_dtype(79),
"objectness_score": fp_dtype(0.77),
}
box_logits = ov.parameter([4, 680], fp_dtype, "box_logits")
class_preds = ov.parameter([4, 170], fp_dtype, "class_preds")
proposals = ov.parameter([4, 1, 8], fp_dtype, "proposals")
aux_class_preds = ov.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ov.parameter([4, 680], fp_dtype, "aux_box_preds")
node = ov.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_num_classes() == int_dtype(85)
assert node.get_background_label_id() == int_dtype(13)
assert node.get_top_k() == int_dtype(16)
assert node.get_variance_encoded_in_target()
assert np.all(np.equal(node.get_keep_top_k(), np.array([64, 32, 16, 8], dtype=int_dtype)))
assert node.get_code_type() == "caffe.PriorBoxParameter.CENTER_SIZE"
assert not node.get_share_location()
assert np.isclose(node.get_nms_threshold(), fp_dtype(0.645))
assert np.isclose(node.get_confidence_threshold(), fp_dtype(0.111))
assert node.get_clip_after_nms()
assert not node.get_clip_before_nms()
assert node.get_decrease_label_id()
assert node.get_normalized()
assert node.get_input_height() == int_dtype(86)
assert node.get_input_width() == int_dtype(79)
assert np.isclose(node.get_objectness_score(), fp_dtype(0.77))
assert node.get_num_classes() == int_dtype(85)
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[

37
src/bindings/python/tests_compatibility/test_ngraph/test_create_op.py
View File

@@ -110,7 +110,7 @@ def test_ctc_greedy_decoder(dtype):
(np.float64, np.int64, "i64", "i32", False, False),
(np.float64, np.int64, "i32", "i64", False, False),
(np.float64, np.int64, "i64", "i64", False, False)
],)
], )
def test_ctc_greedy_decoder_seq_len(fp_dtype, int_dtype, int_ci, int_sl, merge_repeated, blank_index):
input0_shape = [8, 20, 128]
input1_shape = [8]
@@ -1093,41 +1093,6 @@ def test_prior_box_clustered(int_dtype, fp_dtype):
assert list(node.get_output_shape(0)) == [2, 4332]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_detection_output(int_dtype, fp_dtype):
attributes = {
"num_classes": int_dtype(85),
"keep_top_k": np.array([64], dtype=int_dtype),
"nms_threshold": fp_dtype(0.645),
}
box_logits = ng.parameter([4, 8], fp_dtype, "box_logits")
class_preds = ng.parameter([4, 170], fp_dtype, "class_preds")
proposals = ng.parameter([4, 2, 10], fp_dtype, "proposals")
aux_class_preds = ng.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ng.parameter([4, 8], fp_dtype, "aux_box_preds")
node = ng.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_type_name() == "DetectionOutput"
assert node.get_output_size() == 1
assert list(node.get_output_shape(0)) == [1, 1, 256, 7]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[

111
src/bindings/python/tests_compatibility/test_ngraph/test_detection_output.py Normal file
View File

@@ -0,0 +1,111 @@
# Copyright (C) 2018-2021 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
import ngraph as ng
import numpy as np
import pytest
np_types = [np.float32, np.int32]
integral_np_types = [
np.int8,
np.int16,
np.int32,
np.int64,
np.uint8,
np.uint16,
np.uint32,
np.uint64,
]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_detection_output(int_dtype, fp_dtype):
attributes = {
"keep_top_k": np.array([64], dtype=int_dtype),
"nms_threshold": fp_dtype(0.645),
}
box_logits = ng.parameter([4, 8], fp_dtype, "box_logits")
class_preds = ng.parameter([4, 170], fp_dtype, "class_preds")
proposals = ng.parameter([4, 2, 10], fp_dtype, "proposals")
aux_class_preds = ng.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ng.parameter([4, 8], fp_dtype, "aux_box_preds")
node = ng.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_type_name() == "DetectionOutput"
assert node.get_output_size() == 1
assert list(node.get_output_shape(0)) == [1, 1, 256, 7]
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_dynamic_get_attribute_value(int_dtype, fp_dtype):
attributes = {
"background_label_id": int_dtype(13),
"top_k": int_dtype(16),
"variance_encoded_in_target": True,
"keep_top_k": np.array([64, 32, 16, 8], dtype=int_dtype),
"code_type": "caffe.PriorBoxParameter.CENTER_SIZE",
"share_location": False,
"nms_threshold": fp_dtype(0.645),
"confidence_threshold": fp_dtype(0.111),
"clip_after_nms": True,
"clip_before_nms": False,
"decrease_label_id": True,
"normalized": True,
"input_height": int_dtype(86),
"input_width": int_dtype(79),
"objectness_score": fp_dtype(0.77),
}
box_logits = ng.parameter([4, 680], fp_dtype, "box_logits")
class_preds = ng.parameter([4, 170], fp_dtype, "class_preds")
proposals = ng.parameter([4, 1, 8], fp_dtype, "proposals")
aux_class_preds = ng.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ng.parameter([4, 680], fp_dtype, "aux_box_preds")
node = ng.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_background_label_id() == int_dtype(13)
assert node.get_top_k() == int_dtype(16)
assert node.get_variance_encoded_in_target()
assert np.all(np.equal(node.get_keep_top_k(), np.array([64, 32, 16, 8], dtype=int_dtype)))
assert node.get_code_type() == "caffe.PriorBoxParameter.CENTER_SIZE"
assert not node.get_share_location()
assert np.isclose(node.get_nms_threshold(), fp_dtype(0.645))
assert np.isclose(node.get_confidence_threshold(), fp_dtype(0.111))
assert node.get_clip_after_nms()
assert not node.get_clip_before_nms()
assert node.get_decrease_label_id()
assert node.get_normalized()
assert node.get_input_height() == int_dtype(86)
assert node.get_input_width() == int_dtype(79)
assert np.isclose(node.get_objectness_score(), fp_dtype(0.77))

65
src/bindings/python/tests_compatibility/test_ngraph/test_dyn_attributes.py
View File

@@ -1,11 +1,10 @@
# Copyright (C) 2018-2021 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
import ngraph as ng
import numpy as np
import pytest
import ngraph as ng
@pytest.fixture()
def _proposal_node():
@@ -37,68 +36,6 @@ def test_dynamic_attributes_softmax():
assert node.get_axis() == 3
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[
(np.int8, np.float32),
(np.int16, np.float32),
(np.int32, np.float32),
(np.int64, np.float32),
(np.uint8, np.float32),
(np.uint16, np.float32),
(np.uint32, np.float32),
(np.uint64, np.float32),
(np.int32, np.float16),
(np.int32, np.float64),
],
)
def test_dynamic_get_attribute_value(int_dtype, fp_dtype):
attributes = {
"num_classes": int_dtype(85),
"background_label_id": int_dtype(13),
"top_k": int_dtype(16),
"variance_encoded_in_target": True,
"keep_top_k": np.array([64, 32, 16, 8], dtype=int_dtype),
"code_type": "caffe.PriorBoxParameter.CENTER_SIZE",
"share_location": False,
"nms_threshold": fp_dtype(0.645),
"confidence_threshold": fp_dtype(0.111),
"clip_after_nms": True,
"clip_before_nms": False,
"decrease_label_id": True,
"normalized": True,
"input_height": int_dtype(86),
"input_width": int_dtype(79),
"objectness_score": fp_dtype(0.77),
}
box_logits = ng.parameter([4, 680], fp_dtype, "box_logits")
class_preds = ng.parameter([4, 170], fp_dtype, "class_preds")
proposals = ng.parameter([4, 1, 8], fp_dtype, "proposals")
aux_class_preds = ng.parameter([4, 4], fp_dtype, "aux_class_preds")
aux_box_preds = ng.parameter([4, 680], fp_dtype, "aux_box_preds")
node = ng.detection_output(box_logits, class_preds, proposals, attributes, aux_class_preds, aux_box_preds)
assert node.get_num_classes() == int_dtype(85)
assert node.get_background_label_id() == int_dtype(13)
assert node.get_top_k() == int_dtype(16)
assert node.get_variance_encoded_in_target()
assert np.all(np.equal(node.get_keep_top_k(), np.array([64, 32, 16, 8], dtype=int_dtype)))
assert node.get_code_type() == "caffe.PriorBoxParameter.CENTER_SIZE"
assert not node.get_share_location()
assert np.isclose(node.get_nms_threshold(), fp_dtype(0.645))
assert np.isclose(node.get_confidence_threshold(), fp_dtype(0.111))
assert node.get_clip_after_nms()
assert not node.get_clip_before_nms()
assert node.get_decrease_label_id()
assert node.get_normalized()
assert node.get_input_height() == int_dtype(86)
assert node.get_input_width() == int_dtype(79)
assert np.isclose(node.get_objectness_score(), fp_dtype(0.77))
assert node.get_num_classes() == int_dtype(85)
@pytest.mark.parametrize(
"int_dtype, fp_dtype",
[