[Work In Progress]
Our (my robot and I) interactive painting program uses machine learning and computer vision techniques to create interesting paintings. Ultimately, we aim to create a new aesthetics, or a genre, of art. Of course, robotics in art is not a new concept. Still, I hope that through numerous experiments using NN, CV, RL, and my experiences in oil painting, I can expand possibilities for artistic expression and produce unique and exciting results.
So far, we have trained a neural network to generate brush strokes. Then we used this neural network to train a reinforcement learning agent to generate sequences of actions that control the robot arm to create a final painting that looks like a target image. The user can interact with the program by providing feedback on the painting as it progresses, adjusting the target image, or even painting directly on the canvas. Over the next few months, I plan to update my progress in the project and explore potential applications.
Please feel free to contact me with any questions or feedback. You can also find us on Instagram at yejinkim dot aiart.
