Enhancing the Accuracy in Designing an Image with Style Transfer Learning Method Using Visual Geometry Group (VGG16) over InceptionV3

The main aim of this study is to enhance the accuracy in designing an image with style transfer learning method in novel Visual Geometry Group (VGG-16) over InceptionV3. To perform image style transforming, the novel VGG-16 and InceptionV3 algorithms are utilized. ImageNet's dataset contains 1,281,167 images distributed across 1000 classes, with each image having dimensions of 400 rows and 711 columns. This dataset is used for the image style transfer system. Total iterations of 20 are calculated using ClinCalc keeping G power as 0.8 and α of 0.05. The mean accuracy of the novel VGG-16 is high (70.41%) compared to InceptionV3 (54.40%) algorithm. The significant value of P is 0.0016 (independent-sample t test P < 0.05). This implies that there is a statistically significant difference between the two algorithms based on their accuracy. The mean accuracy of the designing an image with style transfer learning method in novel VGG-16 algorithm is better than the InceptionV3.