The idea of image inpainting is inherited from the ancient technique of manually repairing valuable artworks in an undetectable manner. In this work, when we mention image inpainting, it refers to the techniques of digital image inpainting which have applications in different fields like the repair(s) of historical photographs, removal of unwanted objects or filling in the missing regions in a picture, wiping off visible watermarks etc.

In the recent years, image inpainting has also been adopted in the field of fashion design, marketing, post-processing for movie production or in day-to-day life usage (like posting pictures on the social networking websites).

The different Image Inpainting techniques that are in use currently can be broadly classified into two major categories: Learning based methods and Non-learning based methods. Non-learning based methods majorly include statistical methods or patch-based methods whereas learning based methods include the recent work in this area with the rise in popularity of deep neural networks. Both these categories are discussed in detail in the "State Of The Art" section.

Nowadays, with the rise of the Information Age, the users’ data privacy has become a hot topic of research for the community. The concern around the invasive applications of the image data has piqued the interest of the researchers (and developers) working with the facial recognition tools and technology. Following this line of thought, this work also explores the possibility of using image inpainting techniques to protect the data privacy of the visual content uploaded by the users on social media.