Deep learning-based artistic style transformation algorithm in visual communication

Artistic style transition is a popular visual communication technique, the subject matter of one image with the aesthetics of another to create stunning new works. Deep learning (DL) revolutionized computers by solving algorithm-impractical issues, including voice recognition, machine translation, vision, etc. We collect content and style images for our algorithm in this study. Gaussian filter preprocessing is a technique for improving data. This smoothing method removes visual noise and inconsistencies to let the neural network focus on important information and design aspects. We process preprocessed images using a Dense Convolutional Network (DenseNet) to acquire feature representations. Our technique classifies style and content using Long Short-Term Memory (LSTM) networks.LSTM networks enhance visual communication by effectively capturing and processing sequential visual data, improving the accuracy of tasks like video analysis and image captioning.We can represent the sequential character of creative styles using LSTM networks since they function well with sequence data.These altered images produce an intriguing and creative visual result, which is aesthetically pleasing while retaining the semantic and structural coherence of the source material.We use the Python program recommended for visual communication. The findings showed that in terms of accuracy (96%), precision (93%), and runtime (80%)Compared with other DL-based models we examined, the average performance of our proposed method was greater.This approach successfully generates visually appealing and semantically coherent images, demonstrating the potential of advanced neural networks in creative applications.