Machine, Discourse and Power: From Machine Learning in Construction of 3D Face to Art and Creativity

Abstract

The 3D face reconstruction based on a single image can be done recently by training the machine to learn from the database. The performance and accuracy of 3D facial structure and the geometry have largely improved with the powerful calculation of machine and its expanding data. The 3D face reconstruction can be applied in both still images and moving images and recently such technology draws attention in the art field such as portrait representation, facial recognition, surveillance art, etc. By using artificial intelligence (AI) technology and machine learning, artists are able to create artwork with hybridity characteristic, which consists of both human perspective, and machine creativity, which opens up a new gateway to the art world. The capability of artists in creating artwork, particularly portrait art in this research, has been enhanced in terms of technological and theoretical perspective. Artists can represent faces and understand 3D geometry in new ways by breaking through the traditional limitation; it also provides alternative methods in facial visualization beyond the expectation. Traditional expression and technique can be expended to a new dimension; the technology helps artists to think out of the box and continue to create new movement and ism throughout the art history. Moreover, the traditional belief about the importance of the originality of artists’ mind, sense, perspective and techniques in creating artwork can now be redefined by machine and artificial intelligence. A new form in art and creativity, and even knowledge and history has been developed. This paper attempts to discuss the essential transformation in dimensional imagery and its aesthetic experiences through machine learning means, and the shift of discourse of power from human to machine.