| Autor: |
Mynavathi, R., Rajendran, P., S. P., Karunambika, S., Abishek, S. K., Jagadesh, K., Mafas Ahamed |
| Předmět: |
|
| Zdroj: |
Grenze International Journal of Engineering & Technology (GIJET); 2023, Vol. 9 Issue 2, p626-633, 8p |
| Abstrakt: |
Grayscale image colorization is the process for estimating color values for Grayscale image and video frames to gain more insights of the target images and improve their quality aesthetically. Many colorization techniques have been broadly employed colorizations of old and historical photo, satellite images, scientific image illustrations, medical image processing to improve visual features and appeal. Over a decade, many approaches have been introduced for addressing various colorization issues. Existing methods widely falls into three categories namely scribble-based colorization, transfer-based colorization and Self-supervised colorization. This paper analyses various recent prominent works in the above categories. We analyze the reason for GANs being an effective model for image colorization and finally a recent method proposed for using image colorization and a proxy task using GAN is being reviewed to gain insights from the colorization and transferring the knowledge acquired in downstream tasks like segmentation and classification without manual annotation. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
