cs-sakaryauniversity

Sakarya Üniversitesi Bilgisayar Mühendisliği lisans eğitimim boyunca yapmış olduğum ödevler, projeler, ders notları ve çıkmış sınav soruları

Seam-Carving-Level-Detection-With-CNN

In this study, we made detection of seam carving applied to images, by using deep learning models, 2-dimensional convolutional neural networks (CNN - Convolutional Neural Network), and fully connected (FC - Fully Connected) commonly used in computer vision. The proposed model was developed as a multi-classifier with four outputs consisting of 0%, 10%, 20% and 40% seam carving classes. ImageNet which is an open source image data set is used in the training of created deep learning networks. In addition to the original images, training and testing data sets were obtained using images with reduced dimensions of 10%, 20% and 40% by seam carving in the vertical direction. The performance of the model has been scaled with the confusion matrix and the correct classification rate of the test sets. The accuracy analysis of the model was tested on the images, and the results and accuracy value were shown in detail with confusion, precision, f1-score and recall matrices. For more information: <a href="https://ieeexplore.ieee.org/abstract/document/9254845"><b>Click Here</b></a>

Classification-of-specific-classes-of-cifar-100-dataset

How to classify by using of specific classes of cifar-100?

Laravel_Mail_template

Seam-Carving-Algorithm-With-Python

Seam carving, also known as content-aware image resizing, is one of the methods for adjusting the image size such that the content of the image will be preserved. Image resizing is used to handle different resolutions between modules of various device screens and applications. It can also be used intentionally to corrupt or remove the original content of the image. Therefore, it is of great importance to detect external interference with the images.