rm fluid for custom op

x2paddle/op_mapper/caffe2paddle/caffe_custom_layer/priorbox.py

@@ -13,7 +13,85 @@
 # limitations under the License.
 
 import paddle
-import paddle.fluid as fluid
+from paddle import _C_ops
+from paddle.common_ops_import import Variable, LayerHelper, check_variable_and_dtype, check_type, check_dtype
+
+
+def prior_box(input,
+              image,
+              min_sizes,
+              max_sizes=None,
+              aspect_ratios=[1.],
+              variance=[0.1, 0.1, 0.2, 0.2],
+              flip=False,
+              clip=False,
+              steps=[0.0, 0.0],
+              offset=0.5,
+              min_max_aspect_ratios_order=False,
+              name=None):
+    helper = LayerHelper("prior_box", **locals())
+    dtype = helper.input_dtype()
+    check_variable_and_dtype(
+        input, 'input', ['uint8', 'int8', 'float32', 'float64'], 'prior_box')
+
+    def _is_list_or_tuple_(data):
+        return (isinstance(data, list) or isinstance(data, tuple))
+
+    if not _is_list_or_tuple_(min_sizes):
+        min_sizes = [min_sizes]
+    if not _is_list_or_tuple_(aspect_ratios):
+        aspect_ratios = [aspect_ratios]
+    if not (_is_list_or_tuple_(steps) and len(steps) == 2):
+        raise ValueError('steps should be a list or tuple ',
+                         'with length 2, (step_width, step_height).')
+
+    min_sizes = list(map(float, min_sizes))
+    aspect_ratios = list(map(float, aspect_ratios))
+    steps = list(map(float, steps))
+
+    cur_max_sizes = None
+    if max_sizes is not None and len(max_sizes) > 0 and max_sizes[0] > 0:
+        if not _is_list_or_tuple_(max_sizes):
+            max_sizes = [max_sizes]
+        cur_max_sizes = max_sizes
+
+    if in_dynamic_mode():
+        attrs = ('min_sizes', min_sizes, 'aspect_ratios', aspect_ratios,
+                 'variances', variance, 'flip', flip, 'clip', clip, 'step_w',
+                 steps[0], 'step_h', steps[1], 'offset', offset,
+                 'min_max_aspect_ratios_order', min_max_aspect_ratios_order)
+        if cur_max_sizes is not None:
+            attrs += ('max_sizes', cur_max_sizes)
+        box, var = _C_ops.prior_box(input, image, *attrs)
+        return box, var
+    else:
+        attrs = {
+            'min_sizes': min_sizes,
+            'aspect_ratios': aspect_ratios,
+            'variances': variance,
+            'flip': flip,
+            'clip': clip,
+            'step_w': steps[0],
+            'step_h': steps[1],
+            'offset': offset,
+            'min_max_aspect_ratios_order': min_max_aspect_ratios_order
+        }
+
+        if cur_max_sizes is not None:
+            attrs['max_sizes'] = cur_max_sizes
+
+        box = helper.create_variable_for_type_inference(dtype)
+        var = helper.create_variable_for_type_inference(dtype)
+        helper.append_op(
+            type="prior_box",
+            inputs={"Input": input,
+                    "Image": image},
+            outputs={"Boxes": box,
+                     "Variances": var},
+            attrs=attrs, )
+        box.stop_gradient = True
+        var.stop_gradient = True
+        return box, var
 
 
 class PriorBox(object):
@@ -32,8 +110,7 @@ class PriorBox(object):
         }
 
     def __call__(self, x0, x1):
-        box, var = fluid.layers.prior_box(
-            input=x0, image=x1, **self.priorbox_layer_attrs)
+        box, var = prior_box(input=x0, image=x1, **self.priorbox_layer_attrs)
         box = paddle.reshape(x=box, shape=[1, 1, -1])
         var = paddle.reshape(x=var, shape=[1, 1, -1])
         out = paddle.concat(x=[box, var], axis=1)

x2paddle/op_mapper/caffe2paddle/caffe_custom_layer/roipooling.py

@@ -13,7 +13,50 @@
 # limitations under the License.
 
 import paddle
-import paddle.fluid as fluid
+from paddle import _C_ops
+from paddle import in_dynamic_mode
+from paddle.common_ops_import import Variable, LayerHelper, check_variable_and_dtype, check_type, check_dtype
+
+
+def roi_pool(input,
+             rois,
+             pooled_height,
+             pooled_width,
+             spatial_scale=1.0,
+             rois_num=None,
+             name=None):
+    if in_dynamic_mode():
+        assert rois_num is not None, "rois_num should not be None in dygraph mode."
+        pool_out, argmaxes = _C_ops.roi_pool(
+            input, rois, rois_num, "pooled_height", pooled_height,
+            "pooled_width", pooled_width, "spatial_scale", spatial_scale)
+        return pool_out, argmaxes
+
+    else:
+        check_variable_and_dtype(input, 'input', ['float32'], 'roi_pool')
+        check_variable_and_dtype(rois, 'rois', ['float32'], 'roi_pool')
+        helper = LayerHelper('roi_pool', **locals())
+        dtype = helper.input_dtype()
+        pool_out = helper.create_variable_for_type_inference(dtype)
+        argmaxes = helper.create_variable_for_type_inference(dtype='int32')
+
+        inputs = {
+            "X": input,
+            "ROIs": rois,
+        }
+        if rois_num is not None:
+            inputs['RoisNum'] = rois_num
+        helper.append_op(
+            type="roi_pool",
+            inputs=inputs,
+            outputs={"Out": pool_out,
+                     "Argmax": argmaxes},
+            attrs={
+                "pooled_height": pooled_height,
+                "pooled_width": pooled_width,
+                "spatial_scale": spatial_scale
+            })
+        return pool_out, argmaxes
 
 
 class ROIPooling(object):
@@ -26,6 +69,5 @@ class ROIPooling(object):
 
     def __call__(self, x0, x1):
         slice_x1 = paddle.slice(input=x1, axes=[1], starts=[1], ends=[5])
-        out = fluid.layers.roi_pool(
-            input=x0, rois=slice_x1, **self.roipooling_layer_attrs)
+        out = roi_pool(input=x0, rois=slice_x1, **self.roipooling_layer_attrs)
         return out

x2paddle/op_mapper/onnx2paddle/onnx_custom_layer/nms.py

@@ -13,9 +13,9 @@
 # limitations under the License.
 
 import paddle
-from paddle.fluid import core
-from paddle.fluid.framework import Variable, in_dygraph_mode
-from paddle.fluid.layer_helper import LayerHelper
+from paddle import _C_ops
+from paddle import in_dynamic_mode
+from paddle.common_ops_import import Variable, LayerHelper
 
 
 def multiclass_nms(bboxes,
@@ -33,12 +33,12 @@ def multiclass_nms(bboxes,
                    name=None):
     helper = LayerHelper('multiclass_nms3', **locals())
 
-    if in_dygraph_mode():
+    if in_dynamic_mode():
         attrs = ('background_label', background_label, 'score_threshold',
                  score_threshold, 'nms_top_k', nms_top_k, 'nms_threshold',
                  nms_threshold, 'keep_top_k', keep_top_k, 'nms_eta', nms_eta,
                  'normalized', normalized)
-        output, index, nms_rois_num = core.ops.multiclass_nms3(bboxes, scores,
+        output, index, nms_rois_num = _C_ops.multiclass_nms3(bboxes, scores,
                                                              rois_num, *attrs)
         if not return_index:
             index = None