By Radina Yotova

By the constant improvement of networking and cameras, digital images became a powerful tool for engagement across social media platforms. They are mostly stored in devices, thus making it easily accessible to revisit or distribute them across a variety of social media platforms such as Facebook, Instagram and TikTok. The materiality of the digital image has a liquid nature in that it exists as a binary code and therefore may be easily deconstructed, re-constructed and re-imagined. Social media platforms are capable of analysing every single image that appears on their servers. By using facial recognition algorithms, which are trained to match a human face from a still or moving image against a database of faces, they can easily verify a user.

Identity 2.0 investigates the methods by which social media platforms use facial recognition systems to identify human faces in digital images, and the extent of accuracy to which those algorithms are capable of performing.

For Identity 2.0 Radina Yotova had requested her personal information data from Facebook as a starting point. In order to translate her research visually, she had worked with a selection of all the selfie images she has on her account.

The moving image is a representation of how people’s facial features are being identified by the facial recognition system. It is emphasising the utmost importance of understanding the way our identities are being captured and processed by the technology.

[1] Keep Dean (January 2014) The Liquid Aesthetic of the Cameraphone: Re-imagining Photography in the Mobile Age, Journal of Creative Technologies (MINA Special Issue), 4, 128-146.

[2] Alshawaf Eman (June 2016) iPhoneography and New Aesthetics: The Emergence of a Social Visual Communication Through Image-based Social Media. Available from https://www.researchgate.net/publication/346774014_iPhoneography_and_New_Aesthetics_The_Emergence_of_a_Social_Visual_Communication_Through_Image-based_Social_Medi

[3] Dr. Rubinstein Daniel (June 2005) Cellphone photography; The death of the camera and the arrival of visible speech, Issues in contemporary culture and aesthetics (2005).

[4] Sefik Ilkin Serengil (February 17, 2020) Face Recognition with Facebook DeepFace in Keras. Available from https://sefiks.com/2020/02/17/face-recognition-with-facebook-deepface-in-keras/

Evan Roth, Since You Were Born (2020)

Coralie Vogelaar, Collection of work

Jan Robert Leegte, Collection of work

Sterling Crispin, Data-Masks (2013-2015)

Identity 2.0 is based on a selection of all the selfie images from Radina Yotova's personal Facebook image dataset.

Region of Interests

The concept of Region of Interests is common in the computer-controlled image processing. ROI is an evaluation software, which supports the user to establish roughly the area within which the evaluation process should operate.

DeepFace

DeepFace is a deep learning facial recognition system created by Facebook. It is identifying human faces in digital images. The program was entirely trained on images uploaded by Facebook users.

Identity 2.0

For Identity 2.0 Radina Yotova have been experimenting with reconstructing cut-outs of the facial features of interest to the software from selfies. The final outcome is a collage-made moving-image, which was later on analysed by an object-detection ML model.

Moving-image compositio