In this artistic installation, users are thrown into the shoes of a woman with a simple task: get back home, safely. Shadows lurk behind every corner, though, letting users feel the anxiety and dread countless women face every night.
This project is envisioned as an artistic installation, therefore it is not meant for users to download it on their devices and run themselves. Rather, the authors intend it to be a thought-provoking piece to be placed in a gallery, possibly augmented by the surrounding environment. The users could then interact with it via a VR headset, for better immersion.
The installation revolves around making the user feel uneasy and paranoid, mirroring the experience many women feel whilst alone at night. It is the hope of the authors that such a project could convey said emotions to users who normally don't feel them, be it because of gender or geographical location, and spread awareness of the situation and empathy towards those women.
The exeprience is modeled after a walk back home at night. No instructions are given to the users, other than where it is headed, and it is placed in a random position in the map, to increase the sense of bewilderment. No landmarks are present on the map, also for the same reason, other than a few scattered writings on buildings that provide some data on cases of violence towards women. Although these could breach the immersion, the ultimate goal is to spread awareness, therefore it was deemed beneficial, along with surprising the users with other events whilst they are not expecting them. A few shadowy entities are also placed, to scare the users should they get near them, along with other assorted noises scattered throughout.
To accomplish a higher level of interactivity and immersion, VR was chosen as a technology. The users are therefore able to move freely in a reactive 3D environment, created using Unity, augmented with spacial audio effects and an evolving audio background, powered by a custom synth designed in SuperCollider and controlled with the aid of a recurrent neural network, built using Python. OSC messaging is used to connect the components together. Supporting libraries for OSC messaging in Unity and machine learning in Python were also used.
The main challenge faced was with the usage of Unity, a technology neither of the authors had prior experience with. Sharing and testing code was therefore difficult, since Git faced issues with the large volume of files and had difficult coping with the changes. Developing for VR was also tricky, since only one headset was available and the authors were physically distant.
Creating the RNN also turned out to be problematic, since few data was available to train it for a virtual 3D environment. Ultimately, trial-and-error was used to refine its final output.
Coordination and collaboration on different parts of the project simultaneously was the strategy used, with the authors deciding on dividing the design spaces between each other, whist communicating changes and helping each other with the Issues faced in their respective domains.
A deeper understanding of the main theme was also reached: whilst the authors started with a specific idea, researching data deepened the perspective on such matters.
This project was built by:
- Coccoli Veronica: neural network design and Python programming
- Martelli Mattia: audio design and SuperCollider programming
- Paglialunga Andrea: environment design and Unity programming
- Portentoso Alice: environment design and Unity programming
 |
 |
 |
|---|