Deep Learning Implementation for Image Classification on Dummy Data with Resolution Variations

This study investigates the performance of deep learning models in classifying image data with varying resolutions. By utilizing dummy data, we aim to establish a controlled environment for evaluating the impact of image resolution on classification accuracy. A variety of deep learning architectures, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and their hybrid combinations, are employed. These models are trained on datasets generated with different image resolutions, ranging from low to high. The classification task focuses on distinguishing between predefined categories within the dummy data. To assess the performance of the models, metrics such as accuracy, precision, recall, F1-score, and confusion matrices are utilized. The impact of resolution variations on these metrics is analyzed, providing insights into the sensitivity of deep learning models to image quality.