NO PHOTOS PLEASE

NO PHOTOS PLEASE

"NO PHOTOS PLEASE" is an infinity mirror whose only purpose is for people to look at it, experience it, but above all, not take pictures of it. The mirror uses a webcam and an artificial intelligence model has been trained to recognize when someone is trying to take a picture of it. When it does, it turns off the lights in the infinity mirror, thus converting it into a quotidian one. By doing this, the mirror hopes to nudge people into deleting any pictures they may have taken of it and fully engage with the mirror's humorous and unique aspects in the present moment.


OFF

ON

TRAINING AN ai RECOGNITION MODEL

Upon exploring alternatives, I discovered YOLO models for phone recognition. However, I lacked the time and expertise to train a dedicated YOLO model. Instead, I found a MobileNet model that efficiently identified mobile phones. Regrettably, it required a powerful PC to run smoothly and lagged on my portable computer.

I had no prior experience with Python coding. As a result, I heavily relied on ChatGPT to handle the majority, if not all, of the coding for me. However, throughout this project, I did manage to learn a great deal. The image attached showcases the various scripts I created in my attempts to make object identification work AS WELL AS TRIGGERING RELAY TO TURN THE LIGHT ON AND OFF FROM THE MIRROR.

SHOWN DURING DESIGN WEEK AT AMERICAN DESIGN CLUB

Initially, I wanted to utilize Google's Teachable Machine to train a model capable of recognizing when someone had a phone out. However, I quickly discovered that this endeavor proved to be more challenging than I had initially anticipated. Several factors, including variations in lighting, different angles in different locations, and the model's difficulty in identifying individuals other than myself, presented significant obstacles.

Pressed for time, I returned to Google's Teachable Machine as a last resort. However, this time, I employed a larger dataset that included images of people taking selfies, as well as comparative images of rooms and people standing around. These images encompassed various angles and lighting conditions.

Previous
Previous

WOOD PLASTIC METAL (2023)

Next
Next

CATCH THE WAVE (2022)