MNN/source/backend/nnapi/execution/NNAPIInterp.cpp

48 lines
1.7 KiB
C++

//
// NNAPIInterp.cpp
// MNN
//
// Created by MNN on 2022/09/30.
// Copyright © 2018, Alibaba Group Holding Limited
//
#include "NNAPIInterp.hpp"
namespace MNN {
NNAPIInterp::NNAPIInterp(MNN::Backend *b, const MNN::Op *op, const std::vector<Tensor *> &inputs, const std::vector<MNN::Tensor *> &outputs) : NNAPICommonExecution(b, op) {
}
ErrorCode NNAPIInterp::onResize(const std::vector<Tensor *> &inputs, const std::vector<Tensor *> &outputs) {
auto output = outputs[0];
auto interpParam = mOp->main_as_Interp();
// nearstneighbor/bilinear: [input, ow/sw, oh/sh, NCHW/NHWC, aligncorners, halfpixelcenter]
auto inputIdxs = getTensorIdxs(inputs);
// shape or scale
#if 1
inputIdxs.push_back(buildScalar(output->width()));
inputIdxs.push_back(buildScalar(output->height()));
#else
inputIdxs.push_back(buildScalar(interpParam->widthScale()));
inputIdxs.push_back(buildScalar(interpParam->heightScale()));
#endif
inputIdxs.push_back(buildScalar(mNCHW));
inputIdxs.push_back(buildScalar(interpParam->alignCorners()));
// inputIdxs.push_back(buildScalar(interpParam->halfPixelCenters()));
inputIdxs.push_back(buildScalar(!interpParam->halfPixelCenters()));
int activateType = -1;
if (interpParam->resizeType() == 1) {
activateType = ANEURALNETWORKS_RESIZE_NEAREST_NEIGHBOR;
} else if (interpParam->resizeType() == 2) {
activateType = ANEURALNETWORKS_RESIZE_BILINEAR;
} else {
MNN_ERROR("[NNAPI] Interp Don't support [Cubic, NearestneighborRound] mode.");
return NOT_SUPPORT;
}
return buildOperation(activateType, inputIdxs, getTensorIdxs(outputs));
}
REGISTER_NNAPI_OP_CREATOR(NNAPIInterp, OpType_Interp)
} // namespace MNN