This thesis presents a method to remove rain from the image. It is divided into light rain and heavy rain, which have different methods of removing rain. Because there are two ways to remove the rain. Precipitation separation ...

This thesis presents a method for improving the efficiency of lip reading through the analysis of keyframes using CNN and LSTM working together, which combines the characteristics of image-based learning with sequential ...

This thesis designed and developed a deep learning model to create Thai image captions using Convolutional Neural Network (CNN) such as VGG16 and others to extract image features and use Bidirectional LSTM is used to create ...