/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserve. 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. */ #include "paddle/fluid/operators/interpolate_op.h" #include #include #include "paddle/fluid/framework/op_registry.h" namespace paddle { namespace operators { using framework::Tensor; class InterpolateOp : public framework::OperatorWithKernel { public: using framework::OperatorWithKernel::OperatorWithKernel; protected: void InferShape(framework::InferShapeContext* ctx) const override { PADDLE_ENFORCE(ctx->HasInput("X"), "Input(X) of InterpolateOp should not be null."); PADDLE_ENFORCE(ctx->HasOutput("Out"), "Output(Out) of InterpolationOp should not be null."); auto interp_method = ctx->Attrs().Get("interp_method"); PADDLE_ENFORCE( "bilinear" == interp_method || "nearest" == interp_method, "Interpolation method can only be \"bilinear\" or \"nearest\"."); auto dim_x = ctx->GetInputDim("X"); // NCHW format int out_h = ctx->Attrs().Get("out_h"); int out_w = ctx->Attrs().Get("out_w"); PADDLE_ENFORCE_EQ(dim_x.size(), 4, "X's dimension must be 4"); if (ctx->HasInput("OutSize") && ctx->IsRuntime()) { auto out_size_dim = ctx->GetInputDim("OutSize"); PADDLE_ENFORCE_EQ(out_size_dim.size(), 1, "OutSize's dimension size must be 1"); PADDLE_ENFORCE_EQ(out_size_dim[0], 2, "OutSize's dim[0] must be 2"); ctx->ShareLoD("X", "Out"); return; } std::vector dim_out({dim_x[0], dim_x[1], out_h, out_w}); ctx->SetOutputDim("Out", framework::make_ddim(dim_out)); } protected: framework::OpKernelType GetExpectedKernelType( const framework::ExecutionContext& ctx) const override { return framework::OpKernelType(ctx.Input("X")->type(), ctx.GetPlace()); } }; class InterpolateOpMaker : public framework::OpProtoAndCheckerMaker { public: void Make() override { AddInput("X", "The input tensor of interpolate operator, " "This is a 4-D tensor with shape of [N, C, H, w]."); AddInput("OutSize", "This is a 1-D tensor with two numbers to specify output size. " "The first number is height and the second number is width.") .AsDispensable(); AddOutput("Out", "The output tensor of interpolate operator, " "This is a 4-D tensor with shape of [N, C, H, W]."); AddAttr("out_h", "output height of interpolate op."); AddAttr("out_w", "output width of interpolate op."); AddAttr("interp_method", "(string, default \"bilinear\"), interpolation " "method, can be \"bilinear\" for " "bilinear interpolation and \"nearest\" for nearest " "neighbor interpolation.") .SetDefault("bilinear"); AddComment(R"DOC( This operator samples input X to given output shape by using specified interpolation method, the interpolation methods can be \"nearest\" for nearest neighbor interpolation and \"bilinear\" for bilinear interpolation. Nearest neighbor interpolation is to perform nearest neighbor interpolation in both the 3rd dimention(in height direction) and the 4th dimention(in width direction) on input tensor. Bilinear interpolation is an extension of linear interpolation for interpolating functions of two variables (e.g. H-direction and W-direction in this op) on a rectilinear 2D grid. The key idea is to perform linear interpolation first in one direction, and then again in the other direction. For details of nearest neighbor interpolation, please refer to Wikipedia: https://en.wikipedia.org/wiki/Nearest-neighbor_interpolation For details of bilinear interpolation, please refer to Wikipedia: https://en.wikipedia.org/wiki/Bilinear_interpolation )DOC"); } }; class InterpolateOpGrad : public framework::OperatorWithKernel { public: using framework::OperatorWithKernel::OperatorWithKernel; protected: void InferShape(framework::InferShapeContext* ctx) const override { PADDLE_ENFORCE(ctx->HasInput("X"), "Input(X) should not be null"); PADDLE_ENFORCE(ctx->HasInput(framework::GradVarName("Out")), "Input(Out@GRAD) should not be null"); auto dim_x = ctx->GetInputDim("X"); if (ctx->HasOutput(framework::GradVarName("X"))) { ctx->SetOutputDim(framework::GradVarName("X"), dim_x); } } framework::OpKernelType GetExpectedKernelType( const framework::ExecutionContext& ctx) const override { return framework::OpKernelType(ctx.Input("X")->type(), ctx.GetPlace()); } }; } // namespace operators } // namespace paddle namespace ops = paddle::operators; REGISTER_OPERATOR(bilinear_interp, ops::InterpolateOp, ops::InterpolateOpMaker, paddle::framework::DefaultGradOpDescMaker); REGISTER_OPERATOR(bilinear_interp_grad, ops::InterpolateOpGrad); REGISTER_OPERATOR(nearest_interp, ops::InterpolateOp, ops::InterpolateOpMaker, paddle::framework::DefaultGradOpDescMaker); REGISTER_OPERATOR(nearest_interp_grad, ops::InterpolateOpGrad); REGISTER_OP_CPU_KERNEL(bilinear_interp, ops::InterpolateKernel, ops::InterpolateKernel, ops::InterpolateKernel); REGISTER_OP_CPU_KERNEL(bilinear_interp_grad, ops::InterpolateGradKernel, ops::InterpolateGradKernel); REGISTER_OP_CPU_KERNEL(nearest_interp, ops::InterpolateKernel, ops::InterpolateKernel, ops::InterpolateKernel); REGISTER_OP_CPU_KERNEL(nearest_interp_grad, ops::InterpolateGradKernel, ops::InterpolateGradKernel);