chakradharsrinivas / bringing-old-photos-back-to-life

Bringing-old-photos-Back-to-life

Photographs connect us to our past and remind us of people, feelings but what if these are damaged?

Many advanced Machine Learning Models are developed for restoring images that are damaged.

Bringing old photos back to life is one such project through which one can restore virtually enhanced high-resolution images for the damaged ones.

Although there have been some image restoration procedures proposed in the literature, this problem is far from being solved satisfactorily. Image inpainting is an important task for computer vision applications, due to large modified data using image editing tools. From these applications, we can find image quality enhancement, image restoration,, and others. A brief image inpainting review is performed through the GAN model.

GAN model is proposed and it is applied to inpaint the missing regions. The enhancement is also applied to the image, which in turn results in a high-resolution image. Our project demonstrates good performance in restoring degraded old photos.

The proposed Inpainting may be applied for removing the unwanted object from an image without affecting the background of an image. Extending the project towards the removal of blurring on the borders of coherent inpainted regions through adding extra functions or methods. In conclusion, there is no method that can inpaint all the types of distortion in images, but using learning techniques there are some promising results for each category of analyzed cases.

The more details about the model and running info can be found from the report attached.