Create readme.md

AI Generated Fruits and Real Fruits Classification using Image Processing/Model/readme.md

@@ -0,0 +1,56 @@
+## AI Generated Fruits and Real Fruits Classification using Image Processing
+
+### Goal:
+###### The aim of this project to identify and predict the real fruits and AI generated fruits using Image Processing methods.
+
+### Dataset:
+https://www.kaggle.com/datasets/osmankagankurnaz/dataset-of-ai-generated-fruits-and-real-fruits
+
+### Description:
+###### The project aims to classify a dataset consisting of images of apple (redapple/greenapple) shotted from different angles into Ai generated images or Real images irrespective of the type/color of apple. To the achieve it, images need to be preprocessed and then trained on various models (atleast 3) and then prediction to be done for some input image. Finally accuracy of models are needed to be compared and provide best output.
+
+### What I had done:
+###### As per the problem statement, the project aims to classify whether an image is Real or generated by AI.
+###### The dataset provide for same consisted of 302 images in total.
+###### After loading the dataset, extracting its features, resizing, reshaping and preprocessing it we trained our machine learning and deep learning models and calculated their accuracy and cross verified it by prediction also.
+###### Following are the model we used :
+###### Principle Component Analysis (PCA)
+###### Linear Discriminant Analysis (LDA)
+###### Support Vector Machine (SVM) (All the three kernel viz., linear, poly and RBF)
+###### CNN (VGG16)
+###### CNN (ResNet)
+###### The accuracy percentage of our models was very satisfying (all nearly about 90% to 100% accurate) and also the predictions were accurate.
+
+
+### Models used:
+###### Principle Component Analysis (PCA)
+###### Linear Discriminant Analysis (LDA)
+###### Support Vector Machine (SVM) all three kernels
+###### CNN- VGG16
+###### CNN- ResNet
+
+### Libraries needed:
+###### pandas
+###### numpy
+###### matplotlib
+###### scikit-learn
+###### tensorflow
+###### keras
+###### openCV
+
+### Visualizations:
+##### PCA-graph:
+![image](https://github.com/AgrawalTitiksha/DL-Simplified/assets/117917014/e10db31b-9749-49ad-b38c-2551c96aeb9c)
+
+### Accuracies:
+![image](https://github.com/AgrawalTitiksha/DL-Simplified/assets/117917014/7432d230-138b-4e8f-b81c-93300629c083)
+
+### Conclusion: 
+###### After evaluating the models, all models are giving pretty much same accuracy and all of them are giving the correct prediction. Both the CNN models are being trained to give 100 % accuracy and SVM model is also giving 100% accuracy. The aim is achieved.
+
+### Contributor:
+###### Name: Titiksha Agrawal
+###### linkedin: 
+https://www.linkedin.com/in/titiksha-agrawal-056004251/
+###### github:
+https://github.com/AgrawalTitiksha