[ad_1]
Riding the wave of generative AI, Picsart, developer of a range of photo and video editing apps for the web and mobile devices, is introducing a new iOS app that transforms your photos and drawings into digital art. Called SketchAI, the app allows users to sketch photos or upload existing images to apply different artistic styles.
SketchAI is easy to use. There are several pre-selected styles you can apply to your creations, including ink drawings, pencil sketches, and artist-inspired ‘Da Vinci’ and ‘Van Gogh’. In addition to sketching and uploading photos, users can add prompts that describe the image (such as ‘Aivazovsky pointing at a boat at night’) to enhance the generated results.
SketchAI offers 5 free creations. A subscription of $5.99 per week to $17.99 per month or $69.99 per year is required to unlock unlimited generations.
“We have a lot planned on how to improve user experience, processing and image quality, add more prompts and artistic style,” Picsart VP Lusine Harutyunyan of the product told TechCrunch. said in an email interview. “Picsart is known for its powerful and fun editing tools, so we are also thinking of adding features that allow users to enhance and manipulate their results. We are pleased to offer a more unique entry point for this creative technology.”
SketchAI joins Picsart’s list of generative AI tools, including AI Avatar, which creates custom AI-generated profile pictures from selfies, coming as art-generating AI apps draw controversy from both users and the art community To do. Lensa’s recently launched viral avatar creator came under scrutiny for its bias against the sexual depiction of women. Meanwhile, art portal ArtStation, which began allowing AI art for the first time earlier this year, has its members widely protested by posting “no AI art” images in their portfolios, while AI-generated art is seen as an art form on the platform. claimed to be a threat to integrity.
Image credit: Pisart
Harutyunyan does not deny that there are problems with generative AI systems, acknowledging that the system that drives SketchAI — an open-source model called Stable Diffusion — can reproduce biases in the artwork it creates. rice field. But he argued that SketchAI and generative AI as a whole will evolve and improve over time as more people use the technology and additional models become available.
“These are very early stages of the overall generative AI landscape, and the technology continues to evolve rapidly and continue to adhere to industry standards and best practices. Our goal is to empower creators. It’s about supporting artists all over the world,” said Harutyunyan.
Trained on images from the web, including the art community, Stable Diffusion has spread like wildfire in recent months. Lensa and DeviantArt use it to generate images, as do game developer Latitude and countless others. Most use cases are harmless. However, some groups use Stable Diffusion to create objectionable content such as depictions of violence and pornographic, non-consensual deepfakes of celebrities.
Stability AI was also the subject of a recent critical letter from US Congressman Anna G. Eshu (D-California) to the National Security Advisor (NSA) and the Office of Science and Technology Policy. Addresses the release of “not moderating content created on the platform” and “insecure AI models”.
According to Harutyunyan, Picsart tweaked and optimized Stable Diffusion in SketchAI for “quality, intensity of the original image” and put filters in place to prevent “unsafe” art from being generated. However, Picsart has not established a way for artists who do not want SketchAI users to create art in their own style to “opt out”. In recent months, artists such as Hollie Mengert and Greg Rutkowski (whose names have become some of the most commonly used prompts in Stable Diffusion-powered apps) have shown poor AI imitation. Despite the fact that they are accusing them of being tied to their work.
Image credit: Pisart
Stability AI, the startup primarily funding the development of Stable Diffusion, recently bowed to pressure and announced that artists would be able to opt out of the datasets used to train the next generation of Stable Diffusion models. suggesting. Harutyunyan said he will consider adopting the model into SketchAI once Picsart is released.
“The baseline model we use is trained on data rather than on a specific material, rather than reproducing the work of a specific artist,” Harutyunyan said.
However, a word of warning to those who share their SketchAI creations publicly: Picsart states that it cannot guarantee that users can claim copyright on their work. This is because the copyright situation for AI-generated artwork is somewhat fluid at the moment. The United States Patent and Trademark Office (USPTO) recently moved to revoke copyright protection for AI-generated comics, saying copyrighted works require human authors. An ongoing class action lawsuit through the U.S. court system also claims that generative AI violates intellectual property law by regurgitating some of the copyrighted data used to develop it. claims to be
Copyright issues have rocked platforms such as Kickstarter and Getty Images, both of which have in recent months banned AI-generated art and the tools to create it for fear of legal repercussions. strictly or completely prohibited.
“As between Picsart and you, you own the rights to your content,” Harutyunyan said. “However, users should be aware of the inherent limitations that come with generative AI.”
[ad_2]
Source link
