diff --git a/modules/gpuarithm/include/opencv2/gpuarithm.hpp b/modules/gpuarithm/include/opencv2/gpuarithm.hpp
index 8996372b59..2bf60eff32 100644
--- a/modules/gpuarithm/include/opencv2/gpuarithm.hpp
+++ b/modules/gpuarithm/include/opencv2/gpuarithm.hpp
@@ -170,17 +170,17 @@ CV_EXPORTS void merge(const std::vector<GpuMat>& src, OutputArray dst, Stream& s
 CV_EXPORTS void split(InputArray src, GpuMat* dst, Stream& stream = Stream::Null());
 CV_EXPORTS void split(InputArray src, std::vector<GpuMat>& dst, Stream& stream = Stream::Null());
 
-//! implements generalized matrix product algorithm GEMM from BLAS
-CV_EXPORTS void gemm(const GpuMat& src1, const GpuMat& src2, double alpha,
-    const GpuMat& src3, double beta, GpuMat& dst, int flags = 0, Stream& stream = Stream::Null());
-
 //! transposes the matrix
 //! supports matrix with element size = 1, 4 and 8 bytes (CV_8UC1, CV_8UC4, CV_16UC2, CV_32FC1, etc)
-CV_EXPORTS void transpose(const GpuMat& src1, GpuMat& dst, Stream& stream = Stream::Null());
+CV_EXPORTS void transpose(InputArray src1, OutputArray dst, Stream& stream = Stream::Null());
 
 //! reverses the order of the rows, columns or both in a matrix
 //! supports 1, 3 and 4 channels images with CV_8U, CV_16U, CV_32S or CV_32F depth
-CV_EXPORTS void flip(const GpuMat& a, GpuMat& b, int flipCode, Stream& stream = Stream::Null());
+CV_EXPORTS void flip(InputArray src, OutputArray dst, int flipCode, Stream& stream = Stream::Null());
+
+//! implements generalized matrix product algorithm GEMM from BLAS
+CV_EXPORTS void gemm(const GpuMat& src1, const GpuMat& src2, double alpha,
+    const GpuMat& src3, double beta, GpuMat& dst, int flags = 0, Stream& stream = Stream::Null());
 
 //! transforms 8-bit unsigned integers using lookup table: dst(i)=lut(src(i))
 //! destination array will have the depth type as lut and the same channels number as source
diff --git a/modules/gpuarithm/src/core.cpp b/modules/gpuarithm/src/core.cpp
index 92f085ebc3..c8ef966e51 100644
--- a/modules/gpuarithm/src/core.cpp
+++ b/modules/gpuarithm/src/core.cpp
@@ -53,9 +53,9 @@ void cv::gpu::merge(const std::vector<GpuMat>&, OutputArray, Stream&) { throw_no
 void cv::gpu::split(InputArray, GpuMat*, Stream&) { throw_no_cuda(); }
 void cv::gpu::split(InputArray, std::vector<GpuMat>&, Stream&) { throw_no_cuda(); }
 
-void cv::gpu::transpose(const GpuMat&, GpuMat&, Stream&) { throw_no_cuda(); }
+void cv::gpu::transpose(InputArray, OutputArray, Stream&) { throw_no_cuda(); }
 
-void cv::gpu::flip(const GpuMat&, GpuMat&, int, Stream&) { throw_no_cuda(); }
+void cv::gpu::flip(InputArray, OutputArray, int, Stream&) { throw_no_cuda(); }
 
 void cv::gpu::LUT(const GpuMat&, const Mat&, GpuMat&, Stream&) { throw_no_cuda(); }
 
@@ -182,13 +182,16 @@ namespace arithm
     template <typename T> void transpose(PtrStepSz<T> src, PtrStepSz<T> dst, cudaStream_t stream);
 }
 
-void cv::gpu::transpose(const GpuMat& src, GpuMat& dst, Stream& s)
+void cv::gpu::transpose(InputArray _src, OutputArray _dst, Stream& _stream)
 {
+    GpuMat src = _src.getGpuMat();
+
     CV_Assert( src.elemSize() == 1 || src.elemSize() == 4 || src.elemSize() == 8 );
 
-    dst.create( src.cols, src.rows, src.type() );
+    _dst.create( src.cols, src.rows, src.type() );
+    GpuMat dst = _dst.getGpuMat();
 
-    cudaStream_t stream = StreamAccessor::getStream(s);
+    cudaStream_t stream = StreamAccessor::getStream(_stream);
 
     if (src.elemSize() == 1)
     {
@@ -260,7 +263,7 @@ namespace
     };
 }
 
-void cv::gpu::flip(const GpuMat& src, GpuMat& dst, int flipCode, Stream& stream)
+void cv::gpu::flip(InputArray _src, OutputArray _dst, int flipCode, Stream& stream)
 {
     typedef void (*func_t)(const GpuMat& src, GpuMat& dst, int flipCode, cudaStream_t stream);
     static const func_t funcs[6][4] =
@@ -273,10 +276,13 @@ void cv::gpu::flip(const GpuMat& src, GpuMat& dst, int flipCode, Stream& stream)
         {NppMirror<CV_32F, nppiMirror_32f_C1R>::call, 0, NppMirror<CV_32F, nppiMirror_32f_C3R>::call, NppMirror<CV_32F, nppiMirror_32f_C4R>::call}
     };
 
+    GpuMat src = _src.getGpuMat();
+
     CV_Assert(src.depth() == CV_8U || src.depth() == CV_16U || src.depth() == CV_32S || src.depth() == CV_32F);
     CV_Assert(src.channels() == 1 || src.channels() == 3 || src.channels() == 4);
 
-    dst.create(src.size(), src.type());
+    _dst.create(src.size(), src.type());
+    GpuMat dst = _dst.getGpuMat();
 
     funcs[src.depth()][src.channels() - 1](src, dst, flipCode, StreamAccessor::getStream(stream));
 }
diff --git a/samples/gpu/driver_api_multi.cpp b/samples/gpu/driver_api_multi.cpp
index 8b4623f41b..a5343df4e9 100644
--- a/samples/gpu/driver_api_multi.cpp
+++ b/samples/gpu/driver_api_multi.cpp
@@ -130,12 +130,12 @@ void Worker::operator()(int device_id) const
     rng.fill(src, RNG::UNIFORM, 0, 1);
 
     // CPU works
-    transpose(src, dst);
+    cv::transpose(src, dst);
 
     // GPU works
     GpuMat d_src(src);
     GpuMat d_dst;
-    transpose(d_src, d_dst);
+    gpu::transpose(d_src, d_dst);
 
     // Check results
     bool passed = norm(dst - Mat(d_dst), NORM_INF) < 1e-3;
diff --git a/samples/gpu/multi.cpp b/samples/gpu/multi.cpp
index 34b111829c..0e9bef6368 100644
--- a/samples/gpu/multi.cpp
+++ b/samples/gpu/multi.cpp
@@ -87,12 +87,12 @@ void Worker::operator()(int device_id) const
     rng.fill(src, RNG::UNIFORM, 0, 1);
 
     // CPU works
-    transpose(src, dst);
+    cv::transpose(src, dst);
 
     // GPU works
     GpuMat d_src(src);
     GpuMat d_dst;
-    transpose(d_src, d_dst);
+    gpu::transpose(d_src, d_dst);
 
     // Check results
     bool passed = norm(dst - Mat(d_dst), NORM_INF) < 1e-3;