Lenet operators

Added backend_cuda for the following operators:

• AvgPooling
• FC/Gemm
• MaxPooling
• ReLU

include/aidge/backend/cuda/operator/AvgPoolingImpl.hpp

+/********************************************************************************
+ * Copyright (c) 2023 CEA-List
+ *
+ * This program and the accompanying materials are made available under the
+ * terms of the Eclipse Public License 2.0 which is available at
+ * http://www.eclipse.org/legal/epl-2.0.
+ *
+ * SPDX-License-Identifier: EPL-2.0
+ *
+ ********************************************************************************/
+
+#ifndef AIDGE_BACKEND_CUDA_OPERATOR_AVGPOOLINGIMPL_H_
+#define AIDGE_BACKEND_CUDA_OPERATOR_AVGPOOLINGIMPL_H_
+
+#include <array>
+#include <memory>
+#include <tuple>
+#include <vector>
+
+#include <cudnn.h>
+
+#include "aidge/backend/OperatorImpl.hpp"
+#include "aidge/operator/AvgPooling.hpp"
+#include "aidge/utils/Registrar.hpp"
+#include "aidge/utils/Types.h"
+
+#include "aidge/backend/cuda/utils/CudaUtils.hpp"
+
+namespace Aidge {
+template <DimIdx_t DIM>
+class AvgPoolingImpl_cuda : public OperatorImpl {
+private:
+    // CuDNN specific variables
+    cudnnPoolingDescriptor_t mAvgPoolingDesc = nullptr;
+    cudnnPoolingMode_t mMode = CUDNN_POOLING_AVERAGE_COUNT_EXCLUDE_PADDING;
+    std::shared_ptr<Tensor> mInputFallback;
+
+public:
+    AvgPoolingImpl_cuda(const AvgPooling_Op<DIM> &op) : OperatorImpl(op, "cuda") {}
+
+    static std::unique_ptr<AvgPoolingImpl_cuda> create(const AvgPooling_Op<2> &op) {
+        return std::make_unique<AvgPoolingImpl_cuda>(op);
+    }
+
+public:
+    void forward();
+    ~AvgPoolingImpl_cuda();
+
+private:
+    template <class T> void forward_(const Tensor& input);
+};
+
+namespace {
+// add cuda backend to AvgPooling_Op<2> implementation registry
+static Registrar<AvgPooling_Op<2>> registrarAvgPoolingImpl_cuda("cuda", Aidge::AvgPoolingImpl_cuda<2>::create);
+}  // namespace
+}  // namespace Aidge
+
+#endif /* AIDGE_BACKEND_CUDA_OPERATOR_AVGPOOLINGIMPL_H_ */