In the recent years in the Chinese speaking world (China, Hong Kong, and Taiwan), there has been a trend of boys and young men dressing up as girls and young women. They are known as Weiniang (fake girls). Arguably this trend began with the youth sub-culture of Cosplay (Costume Play) tightly associated with the ACG (Anime, Comic and Game) culture. Cosplay has provided a safe haven for meetings of fans of ACG to enter a role-playing world where they adorn clothing to emulate the appearance of any ACG character they take great interest in. Some cosplayers take advantage of this chance to practice a gender ‘swapping’, that is, female cosplayers adopt and perform a male ACG role, or vice versa.
The notion of Weiniang was sprouted from an ACG background and now it has been gaining popularity online and off line. A semi-professional Weiniang performance troupe Alisi Weiniang Tuan was established in 2009 to further promote the notion to make it more viable to the general public. Numerous amateur groups and cosplayers participate in the practice of ‘gender swapping’ and share their footage/photos on social media platforms, in an attempt to be praised by the general public as ‘he is more feminine than a real female!’
This paper aims at examining this popular sub-cultural phenomena as a case study from the aspect of attempting to answer: what types of features of the appearance do fake girls in the Chinese speaking world usually pursue in order to be an ultimate fake girl of ‘being more feminine than real female’? To answer this question, the notion of neural network will be applied (either using Python or Mathematica) to perform facial feature recognition by feeding a large data set of images of fake girls collected from famous fake girl websites of the Chinese speaking world, in the hope to discover whether there are a certain types of appearance and facial features one should adopt in order to be a ‘successful’ fake girl. The answer to the research question would shed a different light on the gender studies in the Chinese speaking world.