• Vedant Fuley
• V Jogendra Rao
• Pushpalata Verma

The technique of turning grayscale photos into colored ones is known as image colorization, and it has several uses, such as enhancing visual media and restoring old photographs. Regression loss functions or classification loss functions can be used by an automatic colorization method to accomplish this transition. The classification loss function, on the other hand, might cause color overflow and necessitate substantial calculation for color categories and balance weights of the ground truth, whilst the regression loss function frequently produces brownish colors. This study uses a deep learning model built on the U-Net architecture to suggest a sophisticated method for image colorization. Our goal is to get beyond the drawbacks of conventional techniques by training the model on a sizable dataset of paired color and greyscale images. Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) are used to assess the model's performance; it received scores of 24.7595 and 0.9260, respectively. When compared to traditional algorithms, these results show a notable improvement in the quality and realism of colorized photographs. This development demonstrates how deep learning methods may produce high-quality colorization, opening the door for further advanced applications across a range of domains.

Image Colorization, Regression Loss, Classification Loss, U-Net Architecture, GAN (Generative Adversarial Networks).

"Historical Colorification using U-Net and CGAN", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 11, page no.g467-g473, November-2024, Available :http://www.jetir.org/papers/JETIR2411644.pdf

"Historical Colorification using U-Net and CGAN", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 11, page no. ppg467-g473, November-2024, Available at : http://www.jetir.org/papers/JETIR2411644.pdf