!1178 Complete vm ops for ROIAlign and ROIAlignGrad

Merge pull request !1178 from zhouneng/add_vm_support_roialign

mindspore/ccsrc/kernel/tbe/tbe_adapter.cc

@@ -82,7 +82,9 @@ static std::map<string, string> tbe_func_adapter_map = {
   {"batch_to_space", "batch_to_space_d"},
   {"resize_bilinear", "resize_bilinear_v2_d"},
   {"resize_bilinear_grad", "resize_bilinear_v2_grad"},
-  {"adam", "apply_adam_d"}};
+  {"adam", "apply_adam_d"},
+  {"r_oi_align", "roi_align"},
+  {"r_oi_align_grad", "roi_align_grad"}};
 
 void TbeAdapter::NormalizeFuncName(std::string *func_name) {
   if (func_name == nullptr) {

mindspore/ops/_op_impl/tbe/__init__.py

@@ -169,3 +169,5 @@ from .log1p import _log1p_tbe
 from .resize_bilinear import _resize_bilinear_tbe
 from .resize_bilinear_grad import _resize_bilinear_grad_tbe
 from .flatten import _flatten_tbe
+from .roi_align import _roi_align_tbe
+from .roi_align_grad import _roi_align_grad_tbe

mindspore/ops/_op_impl/tbe/roi_align.py

+# Copyright 2020 Huawei Technologies Co., Ltd
+#
+# 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.
+# ============================================================================
+
+"""ROIAlign op"""
+from mindspore.ops.op_info_register import op_info_register, TBERegOp, DataType
+
+roi_align_op_info = TBERegOp("ROIAlign") \
+    .fusion_type("OPAQUE") \
+    .async_flag(False) \
+    .binfile_name("roi_align.so") \
+    .compute_cost(10) \
+    .kernel_name("roi_align") \
+    .partial_flag(True) \
+    .attr("spatial_scale", "required", "float", "all") \
+    .attr("pooled_height", "required", "int", "all") \
+    .attr("pooled_width", "required", "int", "all") \
+    .attr("sample_num", "optional", "int", "all", "2") \
+    .attr("roi_end_mode", "optional", "0,1", "1") \
+    .input(0, "features", False, "required", "all") \
+    .input(1, "rois", False, "required", "all") \
+    .input(2, "rois_n", False, "optional", "all") \
+    .output(0, "y", False, "required", "all") \
+    .dtype_format(DataType.F16_5HD, DataType.F16_Default, DataType.I32_Default, DataType.F16_5HD) \
+    .dtype_format(DataType.F32_5HD, DataType.F32_Default, DataType.I32_Default, DataType.F32_5HD) \
+    .get_op_info()
+
+
+@op_info_register(roi_align_op_info)
+def _roi_align_tbe():
+    """ROIAlign TBE register"""
+    return

mindspore/ops/_op_impl/tbe/roi_align_grad.py

+# Copyright 2020 Huawei Technologies Co., Ltd
+#
+# 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.
+# ============================================================================
+
+"""ROIAlignGrad op"""
+from mindspore.ops.op_info_register import op_info_register, TBERegOp, DataType
+
+roi_align_grad_op_info = TBERegOp("ROIAlignGrad") \
+    .fusion_type("OPAQUE") \
+    .async_flag(False) \
+    .binfile_name("roi_align_grad.so") \
+    .compute_cost(10) \
+    .kernel_name("roi_align_grad") \
+    .partial_flag(True) \
+    .attr("xdiff_shape", "required", "listInt", "all") \
+    .attr("pooled_width", "required", "int", "all") \
+    .attr("pooled_height", "required", "int", "all") \
+    .attr("spatial_scale", "required", "float", "all") \
+    .attr("sample_num", "optional", "int", "all") \
+    .input(0, "ydiff", False, "required", "all") \
+    .input(1, "rois", False, "required", "all") \
+    .input(2, "rois_n", False, "optional", "all") \
+    .output(0, "xdiff", False, "required", "all") \
+    .dtype_format(DataType.F32_5HD, DataType.F32_Default, DataType.I32_Default, DataType.F32_5HD) \
+    .get_op_info()
+
+
+@op_info_register(roi_align_grad_op_info)
+def _roi_align_grad_tbe():
+    """ROIAlignGrad TBE register"""
+    return

tests/ut/python/ops/test_ops.py

@@ -942,6 +942,15 @@ test_case_nn_ops = [
         'desc_inputs': [Tensor([[[[1, 2, 3, 4, 5]]]], mstype.float32), Tensor([[[[1, 2, 3, 4, 5]]]], mstype.float32)],
         'desc_bprop': [Tensor([[[[1, 2, 3, 4, 5]]]], mstype.float32)],
         'skip': ['backward']}),
+    ('ROIAlign', {
+        'block': P.ROIAlign(7, 7, 0.03125, 2),
+        'desc_inputs': [[2, 256, 192, 320], [1024, 5]],
+        'desc_bprop': [[7,7]]}),
+    ('ROIAlignGrad', {
+        'block': G.ROIAlignGrad((1, 1, 1, 1), 2, 2, 0.5, 2),
+        'desc_inputs': [[1, 1, 2, 2], [1, 5]],
+        'desc_bprop': [[1, 1, 2, 2]],
+        'skip': ['backward']}),
 ]
 
 test_case_array_ops = [