The Eyes of AI
Project details
- Year
- 2024
- Programme
- Bachelor – Graphic Design
- Practices
- Commercial Practices
- Minor
- Contexts & Situatedness
The Eyes of AI: A Visual Investigation Into Camera Surveillance
The Eyes of AI is a series of frame-by-frame animations derived from livestream camera surveillance footage. This work examines the pervasive and often unnoticed presence of surveillance in our daily lives. The animations serve as visual evidence to this widespread presence.
In our ever-expanding digital world, Artificial Intelligence increasingly integrates into various aspects of our lives, including surveillance. When it comes to visual analysis, how far has AI progressed and has it advanced to the point where it can match or even exceed the human eye?
By exploiting AI as a visual tool, these animations bring the intersection of technology and surveillance to light. They challenge us to reconsider our surroundings and the unseen watchers within them: next time you take a dip in the pool, or cross the road to work, you might want to look around to see who else is watching along.
Surveillance and Artificial Intelligence:
During the process of this research project, I experimented with different AI-based tools to discover the link between surveillance and AI.
The final project is a series of frame-by-frame animations consisting of frames extracted from livestream camera surveillance footage. Each frame is different to the other, but the concept was for each frame to contain a reoccurring element, a person in red functions as the visual metaphor – the red thread – throughout the videos.
For this, I partnered with Daniël Jrifat who developed ‘Red Detection Extractor’, an Ai-based application designed to extract frames of people wearing red clothing. This was supposed to make the process of selecting suitable frames for the animations easier. However, as results and conclusion of this experiment, Ai appears to be just like humans – and easily makes mistakes. These “mistakes” or “false positives” are included in the final outcome of this project.
Additionally, I experimented with Video-LLaVA, an open-source AI tool that provides text-based visual descriptions based on videos and photos, we can compare our own human analysis to that of AI. Do we see the same things as AI? The same details? Is AI smarter than us? Or does AI get it all wrong? In the installation showcased during the Graduation Show, you can compare your own answers to that of AI!
Process of The Eyes of AI
During the process development of this project, I did not only experiment with different AI tools, but also with different animation based methods to discover how I could best translate my views on surveillance into a visual work of art.
Curious to find out more on this project or want to get in touch with me? You can contact me here! pipvangelderen.com or email me: pipvan.g@gmail.com