A new way to create realistic 3D shapes using generative AI
Reports and Proceedings
Updates every hour. Last Updated: 22-Apr-2025 10:08 ET (22-Apr-2025 14:08 GMT/UTC)
A new AI method enables the generation of sharp, high-quality 3D shapes that are closer to the quality of the best 2D image models. Previous approaches typically generated blurry or cartoonish 3D shapes.
The first-known direct observations of the movement of microfiber plastics through a thin layer of soil-like particles show that they tend to tumble, roll and sometimes get stuck in spaces. The findings, reported in the journal, Water Resources Research, mean that the fibers could get easily trapped in sediment. The work helps to improve understanding of the exposure risks and possible health impacts of the pervasive pieces of plastic, which are the largest pollutant in the world by mass.
A group of scientists, including researchers from Rutgers University-New Brunswick, Princeton University and Carleton University, has questioned the accuracy and utility of the metaphor “tipping point” in calling attention to the threat of climate change. The phrase, while perhaps initially useful as a clarion call that warns about sudden, drastic changes, may now be confusing the public and impeding action, researchers said.
Researchers developed a fully integrated photonic processor that can perform all the key computations of a deep neural network on a photonic chip, using light. This advance could improve the speed and energy-efficiency of running intensive deep learning models for demanding applications like lidar, astronomical research, and navigation.
New research co-led by the University at Buffalo proposes thinking both big and small to confirm the existence of primordial black holes, suggesting that their signatures could range from very large — hollow planetoids in space — to minute — microscopic tunnels in everyday materials found on Earth, like rocks, metal and glass.
Researchers have demonstrated a new technique for self-assembling electronic devices. The proof-of-concept work was used to create diodes and transistors, and paves the way for self-assembling more complex electronic devices without relying on existing computer chip manufacturing techniques.