Image Colorization using Deep Learning

preview_player
Показать описание
Part of the ECE 542 Virtual Symposium (Spring 2020)

There are several solutions available for the Image Colorization problem. A lot of research is going on in this field to obtain colorized images that are as realistic as possible. Color of objects such as Cars, Clothes etc. that are predicted can be different from the ground truth but can still be considered accurate. However, the challenge lies in accurately colorizing standard components like skin-tone, eyes, hair, nature, sky, sea, etc. Our goal with this project is to produce colored images from grayscale ones that seem natural to the human eye. In our implementation, we propose a method for image colorization. We preprocess colored images to create grey-scale images which are used as input for our model. We then train our model with these grey-scale images as input and the original colored images as output. Upon training the model, we would be able to use it to generate an RGB image with a reasonable understanding of color within its inherent texture.
Рекомендации по теме
Комментарии
Автор

Very interesting project is this can I get source code it will helpful for me.. plz 🙏

anvianvi
Автор

Can we get source code please? we are also working on a same kinda project

sahityakandru
Автор

Its very interesting project, can i get your source code ? i want to make sone improvements in the project.

eisenergy