SimpleCNN: A Simple Convolutional Neural Network Library

Convolutional Neural Network (CNN) Library in C++ SimpleCNN is a C++ library for building and training convolutional neural networks (CNNs) with ease. It provides a simple interface to create and train CNN models for various tasks, such as image classification.

Getting Started

Prerequisites

• C++ compiler
• OpenCV library (Optional); Only for reading data.

Installation

1. Clone the SimpleCNN repository to your local machine:

   git clone https://github.com/yourusername/SimpleCNN.git
   
   

Build the library using your preferred C++ build system Include the necessary headers and link against the SimpleCNN library in your C++ project.

Example Usage

Here's a simple example of how to use SimpleCNN to create and train a CNN model for image classification:

#include <opencv2/opencv.hpp>
#include <opencv2/core.hpp>
#include <iostream>

#define _CUDA_GPU_ // // Enable GPU usage.
#include "SimpleCNN.h"

#ifdef _CUDA_GPU_
#include "kernel.cuh"
#endif //__CUDACC__

using namespace cv;
using namespace std;
using namespace img_read;
using namespace cnn;

int main(void) {
    // Initialize data and labels
    Directory* dirent = new Directory("path/to/image/folder", img_read::FILENAME_EXTENSION::JPG);
    FileRead* label = new FileRead("path/to/label/file.csv");
    
    // Create a CNN model
    CNN* cnn = new CNN(dirent->getImageSet(), label->getLabel());
    
    // Add layers to the model
    cnn->add(new Conv(3, 3, 1, 1, ACTIVATION::TanH));
    cnn->add(new Pooling(POOLING::Max));
    cnn->add(new Padding(1));
    cnn->add(new Conv(3, 3, 1, 1, ACTIVATION::ReLU));
    cnn->add(new Pooling(POOLING::Average));
    cnn->add(new Padding(1));
    cnn->add(new Pooling(POOLING::Min));
    cnn->add(new Flatten());
    cnn->add(new FullyConnected(256, ACTIVATION::Maxout));
    cnn->add(new FullyConnected(128, ACTIVATION::Sigmoid));
    cnn->add(new FullyConnected(64, ACTIVATION::ReLU));
    cnn->add(new FullyConnected(10, ACTIVATION::Softmax));
    // Add more layers as needed
    
    // Compile the model
    cnn->compile(OPTIMIZER::Mini_SGD, LOSS::CategoricalCrossentropy);
    
    // Train the model
    cnn->fit(epochs, batch_size);
    
    // Make predictions
    cnn->predict(image, true_label);

    cnn->accuracy();
    
    return 0;
}

Please replace "path/to/image/folder" and "path/to/label/file.csv" with the actual paths to your image dataset and label file.