Our project allows the user to create colorful paint splatters, effectively painting the world around them. Open CV and Python is used to track the movement of the user's hands.
We were inspired by the "DDU-DU-DDU-DU" k-pop music video by Black Pink and an interest in using computer vision. Also art and computer science is a rare combination to find in any program, so we wanted to explore that.
Using @victordibia 's hand detection software, we were able to use the camera to detect and track the hand so that they would be able to paint when their palm was open. The splatters of paint fade after a set amount of time.
We examined many samples of example code and pages of documentation in order to gain a better understanding of machine learning before tackling our problem.
It was difficult to come up with a project we were all excited about working on, so we wasted quite a bit of time on Friday night debating what to do. Then it was difficult to find existing code of a neural network that had been trained to recognize hands in different sizes and positions that we could use. Finally, the neural network we were using still misidentified/didn’t detect hands at times. We also weren’t sure what the best way to detect a palm opening was.
Our team is proud of having been able to successfully implement a machine learning feature in our project to mesh the worlds of art and computer programing.
While building this, we learned about different methods of applying machine learning and the importance of proper version control. Also that 36 hours is a lot shorter than it seems.
We hope to be able to give the user more control over the brush's actions. Our goal is to create an AR coloring book as the final product.
We built it with @victordibia 's hand detection software that utilizes TensorFlow, Python's Photo Imaging Library, and a built-in webcam.
Arteck HB030 Portable Keyboard
Intel® Movidius™ Neural Compute Stick
$100 Amazon Gift Cards
Oculus Go (32 GB)
Misfit Shine 2
Fujifilm Instax Mini 26