Merge pull request #10143 from pengli:ocl4dnn

modules/dnn/include/opencv2/dnn/dnn.hpp

@@ -410,13 +410,13 @@ CV__DNN_EXPERIMENTAL_NS_BEGIN
          *  @param outputName name for layer which output is needed to get
          *  @details If @p outputName is empty, runs forward pass for the whole network.
          */
-        CV_WRAP void forward(std::vector<Mat>& outputBlobs, const String& outputName = String());
+        CV_WRAP void forward(OutputArrayOfArrays outputBlobs, const String& outputName = String());
 
         /** @brief Runs forward pass to compute outputs of layers listed in @p outBlobNames.
          *  @param outputBlobs contains blobs for first outputs of specified layers.
          *  @param outBlobNames names for layers which outputs are needed to get
          */
-        CV_WRAP void forward(std::vector<Mat>& outputBlobs,
+        CV_WRAP void forward(OutputArrayOfArrays outputBlobs,
                              const std::vector<String>& outBlobNames);
 
         /** @brief Runs forward pass to compute outputs of layers listed in @p outBlobNames.

modules/dnn/src/dnn.cpp

@@ -1638,7 +1638,7 @@ struct Net::Impl
             CV_Error(Error::StsOutOfRange, "Layer \"" + ld.name + "\" produce only " + toString(ld.outputBlobs.size()) +
                                            " outputs, the #" + toString(pin.oid) + " was requsted");
         }
-        if (preferableBackend != DNN_TARGET_CPU)
+        if (preferableBackend != DNN_BACKEND_DEFAULT)
         {
             // Transfer data to CPU if it's require.
             ld.outputBlobsWrappers[pin.oid]->copyToHost();
@@ -1654,10 +1654,35 @@ struct Net::Impl
         return ld.outputBlobs[pin.oid];
     }
 
+    void getBlob(UMat& umat, const LayerPin& pin)
+    {
+        CV_TRACE_FUNCTION();
+
+        if (!pin.valid())
+            CV_Error(Error::StsObjectNotFound, "Requested blob not found");
+
+        LayerData &ld = layers[pin.lid];
+        if ((size_t)pin.oid >= ld.outputBlobs.size())
+        {
+            CV_Error(Error::StsOutOfRange, "Layer \"" + ld.name + "\" produce only " + toString(ld.outputBlobs.size()) +
+                                           " outputs, the #" + toString(pin.oid) + " was requsted");
+        }
+
+        if (ld.umat_outputBlobs.size() > 0 && !ld.umat_outputBlobs[pin.oid].empty())
+            umat = ld.umat_outputBlobs[pin.oid];
+        else
+            umat = UMat();
+    }
+
     Mat getBlob(String outputName)
     {
         return getBlob(getPinByAlias(outputName));
     }
+
+    void getBlob(UMat& umat, String outputName)
+    {
+        getBlob(umat, getPinByAlias(outputName));
+    }
 };
 
 Net::Net() : impl(new Net::Impl)
@@ -1735,7 +1760,7 @@ Mat Net::forward(const String& outputName)
     return impl->getBlob(layerName);
 }
 
-void Net::forward(std::vector<Mat>& outputBlobs, const String& outputName)
+void Net::forward(OutputArrayOfArrays outputBlobs, const String& outputName)
 {
     CV_TRACE_FUNCTION();
 
@@ -1751,16 +1776,40 @@ void Net::forward(std::vector<Mat>& outputBlobs, const String& outputName)
     LayerPin pin = impl->getPinByAlias(layerName);
     LayerData &ld = impl->layers[pin.lid];
 
-    if (ld.umat_outputBlobs.size() > 0)
+    if (outputBlobs.isUMat())
     {
-        for (int i = 0; i < ld.umat_outputBlobs.size(); i++)
-            ld.umat_outputBlobs[i].copyTo(ld.outputBlobs[i]);
+        if (ld.umat_outputBlobs.size() > 0)
+        {
+            UMat umat;
+            impl->getBlob(umat, layerName);
+            outputBlobs.assign(umat);
+        }
+    }
+    else if (outputBlobs.isMat())
+    {
+        outputBlobs.assign(impl->getBlob(layerName));
+    }
+    else if (outputBlobs.isMatVector())
+    {
+        if (ld.umat_outputBlobs.size() > 0)
+        {
+            for (int i = 0; i < ld.umat_outputBlobs.size(); i++)
+                ld.umat_outputBlobs[i].copyTo(ld.outputBlobs[i]);
+        }
+        std::vector<Mat> & outputvec = *(std::vector<Mat> *)outputBlobs.getObj();
+        outputvec = ld.outputBlobs;
+    }
+    else if (outputBlobs.isUMatVector())
+    {
+        if (ld.umat_outputBlobs.size() > 0)
+        {
+            std::vector<UMat> & outputvec = *(std::vector<UMat> *)outputBlobs.getObj();
+            outputvec = ld.umat_outputBlobs;
+        }
     }
-
-    outputBlobs = ld.outputBlobs;
 }
 
-void Net::forward(std::vector<Mat>& outputBlobs,
+void Net::forward(OutputArrayOfArrays outputBlobs,
                   const std::vector<String>& outBlobNames)
 {
     CV_TRACE_FUNCTION();
@@ -1768,7 +1817,7 @@ void Net::forward(std::vector<Mat>& outputBlobs,
     std::vector<LayerPin> pins;
     for (int i = 0; i < outBlobNames.size(); i++)
     {
-       pins.push_back(impl->getPinByAlias(outBlobNames[i]));
+        pins.push_back(impl->getPinByAlias(outBlobNames[i]));
     }
 
     impl->setUpNet(pins);
@@ -1777,11 +1826,14 @@ void Net::forward(std::vector<Mat>& outputBlobs,
 
     impl->forwardToLayer(impl->getLayerData(out.lid));
 
-    outputBlobs.clear();
+    std::vector<Mat> matvec;
     for (int i = 0; i < pins.size(); i++)
     {
-        outputBlobs.push_back(impl->getBlob(pins[i]));
+        matvec.push_back(impl->getBlob(pins[i]));
     }
+
+    std::vector<Mat> & outputvec = *(std::vector<Mat> *)outputBlobs.getObj();
+    outputvec = matvec;
 }
 
 void Net::forward(std::vector<std::vector<Mat> >& outputBlobs,

modules/dnn/src/layers/fully_connected_layer.cpp

@@ -286,8 +286,13 @@ public:
         UMat biasOnesMat = UMat::ones(outerSize, 1, umat_blobs[0].type());
         for (size_t i = 0; i < inputs.size(); i++)
         {
-            UMat& srcMat = inputs[i];
-            UMat& dstMat = outputs[i];
+            MatShape inshape, outshape;
+            inshape = shape(outerSize, innerSize);
+            outshape = shape(outerSize, numOutput);
+
+            UMat srcMat, dstMat;
+            srcMat = inputs[i].reshape(1, inshape.size(), &inshape[0]);
+            dstMat = outputs[i].reshape(1, outshape.size(), &outshape[0]);
             dstMat.setTo(0.0f);
 
             if (!innerProductOp->Forward(srcMat, umat_blobs[0], (bias) ? umat_blobs[1] : UMat(), dstMat))

modules/dnn/src/ocl4dnn/src/math_functions.cpp

@@ -65,8 +65,6 @@ ocl::Image2D ocl4dnnGEMMCopyBufferToImage(UMat buffer, int offset,
                                           int padded_width, int height,
                                           int width, int ld)
 {
-    ocl::Context ctx = ocl::Context::getDefault();
-    ocl::Queue queue = ocl::Queue::getDefault();
     ocl::Image2D image;
 
     if (!is_matrix_a && transpose)
@@ -192,9 +190,6 @@ static bool ocl4dnnFastImageGEMM(const CBLAS_TRANSPOSE TransA,
     // just padding one line is enough as the sub group block read
     // will clamp to edge according to the spec.
 
-    ocl::Context ctx = ocl::Context::getDefault();
-    ocl::Queue queue = ocl::Queue::getDefault();
-
     ocl::Image2D ImA;
     ocl::Image2D ImB;
 
@@ -446,7 +441,6 @@ bool ocl4dnnGEMV<float>(const CBLAS_TRANSPOSE TransA,
                  const int32_t offx, const float beta, UMat y,
                  const int32_t offy)
 {
-    ocl::Queue queue = ocl::Queue::getDefault();
     bool ret = false;
 
     if (TransA == CblasNoTrans)
@@ -507,8 +501,6 @@ bool ocl4dnnAXPY(const int32_t N, const Dtype alpha,
                  const UMat X, const int32_t offX, UMat Y,
                  const int32_t offY)
 {
-    ocl::Context ctx = ocl::Context::getDefault();
-
     ocl::Kernel oclk_axpy(CL_KERNEL_SELECT("axpy"), cv::ocl::dnn::math_oclsrc);
     if (oclk_axpy.empty())
         return false;

modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp

@@ -198,8 +198,6 @@ void OCL4DNNConvSpatial<Dtype>::collectCommonInformation()
     addDef("as_Dtype2", "as_float2");
     addDef("as_Dtype4", "as_float4");
     addDef("as_Dtype8", "as_float8");
-    addDef("Dtype_ID", (int)CV_32F);
-    addDef("Dtype_SIZE", (int)sizeof(Dtype));
 }
 
 typedef enum {

modules/dnn/src/ocl4dnn/src/ocl4dnn_lrn.cpp

@@ -92,7 +92,6 @@ bool OCL4DNNLRN<Dtype>::Forward(const UMat& bottom, UMat& top)
 template<typename Dtype>
 bool OCL4DNNLRN<Dtype>::crossChannelForward(const UMat& bottom, UMat& top)
 {
-    ocl::Queue queue = ocl::Queue::getDefault();
     CHECK_EQ(phase_test_, true) << "Only support forward inference.";
 
     cl_uint argIdx = 0;

modules/dnn/src/ocl4dnn/src/ocl4dnn_pool.cpp

@@ -97,7 +97,6 @@ bool OCL4DNNPool<Dtype>::Forward(const UMat& bottom,
                                  UMat& top_mask)
 {
     bool ret = true;
-    ocl::Queue queue = ocl::Queue::getDefault();
     size_t global[] = { 128 * 128 };
     size_t local[] = { 128 };
     cl_uint argIdx = 0;

modules/dnn/src/ocl4dnn/src/ocl4dnn_softmax.cpp

@@ -83,7 +83,6 @@ template<typename Dtype>
 bool OCL4DNNSoftmax<Dtype>::Forward(const UMat& bottom, UMat& top)
 {
     bool ret = false;
-    ocl::Queue queue = ocl::Queue::getDefault();
     bool intel_subgroup = ocl::Device::getDefault().intelSubgroupsSupport();
     if (intel_subgroup && inner_num_ < 128)
     {

modules/dnn/src/opencl/conv_layer_spatial.cl

@@ -91,7 +91,6 @@
 #define LOOP(N, VAR, STMT) CAT(LOOP, N)((VAR), (STMT))
 
 #if defined(convolve_simd) || defined(Conv_Interleaved)
-#if Dtype_SIZE == 4
 #define INT_TYPE uint
 #define INT_TYPE2 uint2
 #define INT_TYPE4 uint4
@@ -100,9 +99,6 @@
 #define SUB_GROUP_BLOCK_READ4 intel_sub_group_block_read4
 #define SUB_GROUP_BLOCK_READ8 intel_sub_group_block_read8
 #define SUB_GROUP_BLOCK_READ intel_sub_group_block_read
-#else
-#error "Unsupported type"
-#endif
 #endif
 
 #ifdef KERNEL_BASIC
@@ -186,11 +182,7 @@ __kernel void ConvolveBasic(
 
 #elif defined KERNEL_IDLF
 
-#if TYPE == TYPE_HALF
-#define VLOAD4(_v, _p) do { (_v).s0 = *(_p); (_v).s1 = *(_p + 1); (_v).s2 = *(_p + 2); (_v).s3 = *(_p + 3); } while(0)
-#else
 #define VLOAD4(_v, _p) do { _v = vload4(0, _p); } while(0)
-#endif
 
 // Each work-item computes a OUT_BLOCK_WIDTH * OUT_BLOCK_HEIGHT region of one output map.
 // Each work-group (which will be mapped to 1 SIMD16/SIMD8 EU thread) will compute 16/8 different feature maps, but each feature map is for the same region of the imput image.