in ,

This Bizarre Robot Uses AI to Paint Whatever You Tell It To

With the aid of a robot arm that resembles Dall-E with a paintbrush, AI art transitions from pixels to actual canvases.

Do you remember back then? when small text cues you sent using Dall-E and Stable Diffusion were converted into digital art? That is very 2022.

Introduce yourself to Frida, an AI-driven robot from Carnegie Mellon University who turns your prompts into real paintings complete with expressive brushstrokes using a range of approaches. Most impressively, the bot can alter its trajectory while painting to replicate the iterative process of creating art.

In a video introducing the concept, Peter Scholten brand, a Ph.D. student at CMU’s School of Computer Science and one of the robot’s inventors, said, “It will work with its failures and it will adjust its aims.”

According to the team, Frida intends to investigate the relationship between creativity and robots. The team will present its research paper at the IEEE International Conference on Robotics and Automation in London this May. Robots have previously created art and even displayed it. However, employing the same type of generative intelligence that powers experimental tools like the AI chatbot Chat GPT, Frida is specifically created to cooperate with people.

Framework and Robotics Initiative for Developing Arts, or Frida, is an acronym that also honors the late painter Frida Kahlo. But neither Kahlo nor any other person resemble the off-the-shelf device that was discovered in a cellar. The team’s conviction that Frida is a “robotic painting system, not an artist” is emphasized by the fact that it is currently just a robotic arm with a paintbrush attached.

Frida isn’t coming up with the ideas to communicate, according to Scholten brand. Humans then communicate Frida’s objectives with text inputs. They can even flash photographs they want to see depicted as paintings, or they can show the bot images in a style they want. The researchers are experimenting with audio instructions as well; they had Frida create a picture while playing the excellent song Dancing Queen by ABBA. Humans combine the paint colors in the robot’s palette after Frida suggests them on the screen.

According to Scholten brand, “Frida can assist someone who is encountering physical hurdles to generating visual art.” “Frida can also assist those who lack the time to create art by automating some of the laborious processes. We are currently making progress and collaborating with folks to understand the breadth of Frida’s abilities.”

Although Frida is unlikely to ever achieve the fame of its namesake, its exceptional abilities set it apart from other artistic robots, whose input photos typically match their intended outcomes. The process of creating art is dynamic and ever-evolving, and after plotting its course in a virtual setting, Frida employs machine learning to assess and advance in real time. According to CMU robotics professor James McCann, who makes this observation in the movie, the robot plans “in a space of meaning instead of a space of outputs.”

Like most robots, Frida doesn’t value accuracy and might, for instance, include a “error” like a stray paintbrush in its finished output. Hours are spent on each painting, with frequently whimsical and vibrantly color outcomes. Like most robots, Frida doesn’t value accuracy and might, for instance, include a “error” like a stray paintbrush in its finished output. Hours are spent on each painting, with frequently whimsical and vibrantly color outcomes.

Scholten brand remarked, “There’s this one painting of a frog ballerina that I believe worked out pretty wonderfully. It’s extremely stupid and entertaining, and I thought it was interesting to watch what Frida came up with in response to my input.

To prevent an American or Western bias, the researchers used current news headlines to train their various AI components. They then trained them on visuals and text representative of many cultures.

The potential for AI to create music, lyrics, poetry, and even movie scripts is exciting, but it is also raising questions about ethics and copyright among creatives and legal professionals. Art made by AI isn’t produced in a vacuum. It functions by ingesting and rebuilding previously produced human-made art. Will those real graphic designers, illustrators, composers, and photographers find themselves outsourced as machine-produced art advances and becomes more sophisticated?

Some of the artists I’ve spoken to say they become anxious thinking about the difficult issues that AI art raises. Others are embracing what they see as an inevitable shift, such as Steve Coulson, an enthusiastic comics fan who produced a comics story that was totally illustrated by Mid journey. Coulson has always been a fan of comic book artists, and he believes they “have an eye for dramatic composition and dynamic storytelling that I seriously doubt machine learning will be able to match.” However, it’s a tone of fun to use as a visualization tool for non-artists like myself.

Frida’s inventors share a similar perspective.

In the research article, it is stated that the armed bot is “a robotics project to foster human creativity, rather than replace it, by providing intuitive ways for humans to express their ideas using natural language or example images.”