refine doc, test=develop

paddle/fluid/API.spec

@@ -328,7 +328,7 @@ paddle.fluid.layers.polygon_box_transform (ArgSpec(args=['input', 'name'], varar
 paddle.fluid.layers.yolov3_loss (ArgSpec(args=['x', 'gtbox', 'gtlabel', 'anchors', 'anchor_mask', 'class_num', 'ignore_thresh', 'downsample_ratio', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '991e934c3e09abf0edec7c9c978b4691'))
 paddle.fluid.layers.box_clip (ArgSpec(args=['input', 'im_info', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '397e9e02b451d99c56e20f268fa03f2e'))
 paddle.fluid.layers.multiclass_nms (ArgSpec(args=['bboxes', 'scores', 'score_threshold', 'nms_top_k', 'keep_top_k', 'nms_threshold', 'normalized', 'nms_eta', 'background_label', 'name'], varargs=None, keywords=None, defaults=(0.3, True, 1.0, 0, None)), ('document', 'ca7d1107b6c5d2d6d8221039a220fde0'))
-paddle.fluid.layers.box_decoder_and_assign (ArgSpec(args=['prior_box', 'prior_box_var', 'target_box', 'box_score', 'box_clip', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', 'fb470052db88526a94a7e5de9d9b3a4c'))
+paddle.fluid.layers.box_decoder_and_assign (ArgSpec(args=['prior_box', 'prior_box_var', 'target_box', 'box_score', 'box_clip', 'name'], varargs=None, keywords=None, defaults=(None,)), ('document', '005a5ae47d6c8fff721931d69d072b9f'))
 paddle.fluid.layers.accuracy (ArgSpec(args=['input', 'label', 'k', 'correct', 'total'], varargs=None, keywords=None, defaults=(1, None, None)), ('document', '9808534c12c5e739a10f73ebb0b4eafd'))
 paddle.fluid.layers.auc (ArgSpec(args=['input', 'label', 'curve', 'num_thresholds', 'topk', 'slide_steps'], varargs=None, keywords=None, defaults=('ROC', 4095, 1, 1)), ('document', 'e0e95334fce92d16c2d9db6e7caffc47'))
 paddle.fluid.layers.exponential_decay (ArgSpec(args=['learning_rate', 'decay_steps', 'decay_rate', 'staircase'], varargs=None, keywords=None, defaults=(False,)), ('document', '98a5050bee8522fcea81aa795adaba51'))

paddle/fluid/operators/detection/box_decoder_and_assign_op.cc

@@ -35,8 +35,8 @@ class BoxDecoderAndAssignOp : public framework::OperatorWithKernel {
         ctx->HasInput("BoxScore"),
         "Input(BoxScore) of BoxDecoderAndAssignOp should not be null.");
     PADDLE_ENFORCE(
-        ctx->HasOutput("OutputBox"),
-        "Output(OutputBox) of BoxDecoderAndAssignOp should not be null.");
+        ctx->HasOutput("DecodeBox"),
+        "Output(DecodeBox) of BoxDecoderAndAssignOp should not be null.");
     PADDLE_ENFORCE(
         ctx->HasOutput("OutputAssignBox"),
         "Output(OutputAssignBox) of BoxDecoderAndAssignOp should not be null.");
@@ -68,9 +68,9 @@ class BoxDecoderAndAssignOp : public framework::OperatorWithKernel {
                       "of box_score is [N, classnum], The shape of prior_box "
                       "is [N, 4]");
 
-    ctx->SetOutputDim("OutputBox", framework::make_ddim({target_box_dims[0],
+    ctx->SetOutputDim("DecodeBox", framework::make_ddim({target_box_dims[0],
                                                          target_box_dims[1]}));
-    ctx->ShareLoD("TargetBox", /*->*/ "OutputBox");
+    ctx->ShareLoD("TargetBox", /*->*/ "DecodeBox");
     ctx->SetOutputDim(
         "OutputAssignBox",
         framework::make_ddim({prior_box_dims[0], prior_box_dims[1]}));
@@ -84,38 +84,32 @@ class BoxDecoderAndAssignOpMaker : public framework::OpProtoAndCheckerMaker {
     AddInput(
         "PriorBox",
         "(Tensor, default Tensor<float>) "
-        "Box list PriorBox is a 2-D Tensor with shape [M, 4] holds M boxes, "
-        "each box is represented as [xmin, ymin, xmax, ymax], "
+        "Box list PriorBox is a 2-D Tensor with shape [N, 4] which holds N "
+        "boxes and each box is represented as [xmin, ymin, xmax, ymax], "
         "[xmin, ymin] is the left top coordinate of the anchor box, "
         "if the input is image feature map, they are close to the origin "
         "of the coordinate system. [xmax, ymax] is the right bottom "
         "coordinate of the anchor box.");
     AddInput("PriorBoxVar",
              "(Tensor, default Tensor<float>, optional) "
-             "PriorBoxVar is a 2-D Tensor with shape [M, 4] holds M group "
-             "of variance. PriorBoxVar will set all elements to 1 by "
+             "PriorBoxVar is a 2-D Tensor with shape [N, 4] which holds N "
+             "group of variance. PriorBoxVar will set all elements to 1 by "
              "default.")
         .AsDispensable();
-    AddInput(
-        "TargetBox",
-        "(LoDTensor or Tensor) This input can be a 2-D LoDTensor with shape "
-        "[N, classnum*4]. [N, classnum*4], each box is represented as "
-        "[xmin, ymin, xmax, ymax], [xmin, ymin] is the left top coordinate "
-        "of the box if the input is image feature map, they are close to "
-        "the origin of the coordinate system. [xmax, ymax] is the right "
-        "bottom coordinate of the box. This tensor can contain LoD "
-        "information to represent a batch of inputs. One instance of this "
-        "batch can contain different numbers of entities.");
-    AddInput(
-        "BoxScore",
-        "(LoDTensor or Tensor) This input can be a 2-D LoDTensor with shape "
-        "[N, classnum], each box is represented as [classnum] which is "
-        "the classification probabilities.");
+    AddInput("TargetBox",
+             "(LoDTensor or Tensor) "
+             "This input can be a 2-D LoDTensor with shape "
+             "[N, classnum*4]. It holds N targets for N boxes.");
+    AddInput("BoxScore",
+             "(LoDTensor or Tensor) "
+             "This input can be a 2-D LoDTensor with shape "
+             "[N, classnum], each box is represented as [classnum] which is "
+             "the classification probabilities.");
     AddAttr<float>("box_clip",
                    "(float, default 4.135, np.log(1000. / 16.)) "
                    "clip box to prevent overflowing")
         .SetDefault(4.135f);
-    AddOutput("OutputBox",
+    AddOutput("DecodeBox",
               "(LoDTensor or Tensor) "
               "the output tensor of op with shape [N, classnum * 4] "
               "representing the result of N target boxes decoded with "
@@ -130,12 +124,12 @@ class BoxDecoderAndAssignOpMaker : public framework::OpProtoAndCheckerMaker {
 
 Bounding Box Coder.
 
-Decode the target bounding box with the priorbox information.
+Decode the target bounding box with the prior_box information.
 
-The Decoding schema described below:
+The Decoding schema is described below:
 
     $$
-    oy = (ph \\times pyv \\times ty + py) - \\frac{th}{2} 
+    ox = (pw \\times pxv \\times tx + px) - \\frac{tw}{2} 
     $$
     $$
     oy = (ph \\times pyv \\times ty + py) - \\frac{th}{2}
@@ -149,15 +143,15 @@ The Decoding schema described below:
 
 where `tx`, `ty`, `tw`, `th` denote the target box's center coordinates, width
 and height respectively. Similarly, `px`, `py`, `pw`, `ph` denote the
-priorbox's (anchor) center coordinates, width and height. `pxv`, `pyv`, `pwv`,
-`phv` denote the variance of the priorbox and `ox`, `oy`, `ow`, `oh` denote the
-encoded/decoded coordinates, width and height.
+prior_box's (anchor) center coordinates, width and height. `pxv`, `pyv`, `pwv`,
+`phv` denote the variance of the prior_box and `ox`, `oy`, `ow`, `oh` denote the
+decoded coordinates, width and height in decode_box. 
 
-After box decode, the Assigning schema described below:
+decode_box is obtained after box decode, then assigning schema is described below:
 
-For each priorbox, use the best non-background class's decoded values to 
-updata the priorbox locations and get outputassignbox. So, the shape of
-output_assign_box is the same as priorbox.
+For each prior_box, use the best non-background class's decoded values to 
+update the prior_box locations and get output_assign_box. So, the shape of
+output_assign_box is the same as PriorBox.
 )DOC");
   }
 };

paddle/fluid/operators/detection/box_decoder_and_assign_op.cu

@@ -101,7 +101,7 @@ class BoxDecoderAndAssignCUDAKernel : public framework::OpKernel<T> {
     auto* prior_box_var = context.Input<framework::Tensor>("PriorBoxVar");
     auto* target_box = context.Input<framework::LoDTensor>("TargetBox");
     auto* box_score = context.Input<framework::LoDTensor>("BoxScore");
-    auto* output_box = context.Output<framework::Tensor>("OutputBox");
+    auto* output_box = context.Output<framework::Tensor>("DecodeBox");
     auto* output_assign_box =
         context.Output<framework::Tensor>("OutputAssignBox");
 

paddle/fluid/operators/detection/box_decoder_and_assign_op.h

@@ -27,7 +27,7 @@ class BoxDecoderAndAssignKernel : public framework::OpKernel<T> {
     auto* prior_box_var = context.Input<framework::Tensor>("PriorBoxVar");
     auto* target_box = context.Input<framework::LoDTensor>("TargetBox");
     auto* box_score = context.Input<framework::LoDTensor>("BoxScore");
-    auto* output_box = context.Output<framework::Tensor>("OutputBox");
+    auto* output_box = context.Output<framework::Tensor>("DecodeBox");
     auto* output_assign_box =
         context.Output<framework::Tensor>("OutputAssignBox");
     int roi_num = target_box->dims()[0];

python/paddle/fluid/layers/detection.py

@@ -2238,10 +2238,16 @@ def box_decoder_and_assign(prior_box,
         prior_box_var(${prior_box_var_type}): ${prior_box_var_comment}
         target_box(${target_box_type}): ${target_box_comment}
         box_score(${box_score_type}): ${box_score_comment}
+        box_clip(${box_clip_type}): ${box_clip_comment}
         name(str|None): The name of this operator
     Returns:
-        output_box(${output_box_type}): ${output_box_comment}
-        output_assign_box(${output_assign_box_type}): ${output_assign_box_comment}
+        decode_box(Variable), output_assign_box(Variable):
+
+            two variables:
+
+            - decode_box(${decode_box_type}): ${decode_box_comment}
+            - output_assign_box(${output_assign_box_type}): ${output_assign_box_comment}
+
     Examples:
         .. code-block:: python
 
@@ -2253,13 +2259,13 @@ def box_decoder_and_assign(prior_box,
                 name='target_box', shape=[20, 4*81], dtype='float32')
             scores = fluid.layers.data(
                 name='scores', shape=[20, 81], dtype='float32')
-            output_box, assign_box = fluid.layers.box_decoder_and_assign(
+            decoded_box, output_assign_box = fluid.layers.box_decoder_and_assign(
                 pb, pbv, loc, scores, 4.135)
 
     """
     helper = LayerHelper("box_decoder_and_assign", **locals())
 
-    output_box = helper.create_variable_for_type_inference(
+    decoded_box = helper.create_variable_for_type_inference(
         dtype=prior_box.dtype)
     output_assign_box = helper.create_variable_for_type_inference(
         dtype=prior_box.dtype)
@@ -2274,7 +2280,7 @@ def box_decoder_and_assign(prior_box,
         },
         attrs={"box_clip": box_clip},
         outputs={
-            "OutputBox": output_box,
+            "DecodeBox": decoded_box,
             "OutputAssignBox": output_assign_box
         })
-    return output_box, output_assign_box
+    return decoded_box, output_assign_box

python/paddle/fluid/tests/unittests/test_box_decoder_and_assign_op.py

@@ -87,7 +87,7 @@ class TestBoxDecoderAndAssignOpWithLoD(OpTest):
         }
         self.attrs = {'box_clip': box_clip}
         self.outputs = {
-            'OutputBox': output_box,
+            'DecodeBox': output_box,
             'OutputAssignBox': output_assign_box
         }