If you want to make changes to a public sculpture, you might be able to do so now. Researchers have developed artificial intelligence (AI) technology that can create 3D models of objects on your smartphone using editable 2D photos. You'll be able to make realistic changes to your models and view objects from all angles.
Researchers at Canada's Simon Fraser University (SFU) have developed a new AI technology that allows people to use ordinary photographs to take 3D captures of real-world objects and freely edit their shape and appearance.
similar story
The research was recently presented at the Neural Information Processing Systems Conference (NeurIPS) in New Orleans, USA. Researchers have unveiled a new technique called Proximity Attention Point Rendering (PAPR) that can transform a series of 2D photographs of an object into a 3D point cloud representing the object's shape and appearance, the university said in a press statement.
Each point on the model allows the user to control the object. You can reshape an object by dragging points to change its appearance. 3D models can be viewed from any angle. Users can also turn it into a realistic 2D photo of her showing the edited objects.
“The remarkable success of machine learning in areas such as computer vision and natural language is challenging how traditional 3D graphics pipelines can be modified using the same deep learning-based building blocks that have led to the recent AI runaway. “We encourage researchers to investigate whether it can be redesigned to be more effective,” study author Ke Li said in a statement.
According to Lee, this shows the many possibilities for consumer technology. 3D, he added in a statement, has the potential to become as common as 2D as a medium for visual communication and expression today.
As AI expands rapidly, developers around the world are experimenting with its various capabilities. For example, in July 2023, Stability AI, the company behind Stable Diffusion, one of his popular generative AI tools, published Stable Doodle, which converts doodles into high-quality original images. With the rapid expansion of AI, it is becoming the epicenter of many technological breakthroughs.
