Have a damaged painting? Restore it in just hours with an AI-generated “mask”
Peer-Reviewed Publication
Updates every hour. Last Updated: 22-Jun-2025 23:10 ET (23-Jun-2025 03:10 GMT/UTC)
A new method uses AI to physically restore a damaged painting much more quickly than what’s possible using manual techniques. A digitally generated “mask” in the form of thin film is applied directly to the original painting, and can also be easily removed.
UC Davis researchers have developed a brain-computer interface that can instantaneously translate brain activity into voice as a person tries to speak — effectively creating a digital vocal tract.
A research team at the University of Seville, Spain, developed a novel extension of virus machines, an emerging computing model that draws inspiration from how viruses propagate among hosts. These super virus machines, as the team calls them, address time efficiency limits in basic virus machines. This work was published under the title "Super Virus Machines: Faster Virus Transmission, More Efficiency Using Superchannels" on March 21 in Intelligent Computing, a Science Partner Journal.
Here, researchers from Laval University, Harbin Institute of Technology, University of Toronto, University of L’ Aquila, and University of Rome proposed a frequency multiplexed photothermal correlation tomographic (FM-PCT) technique. This approach overcomes the limitations of 2D imaging modes in infrared thermography, as well as the challenges of low imaging speed, narrow field of view, and low resolution in photothermal imaging.
The leading researcher, Dr Andreas Mandelis, commented: “It is these conditions that, for the first time, make FM-PCT a leading candidate for implementation in fast-turn-around quality control situations, such as industrial manufacturing environments.”
Researchers from Beijing Institute of Technology introduce a novel two-stage method for converting monochromatic near-infrared (NIR) images into high-quality RGB images. In the first stage, luminance information is recovered by converting NIR images into grayscale images. The second stage then restores chrominance information, transforming grayscale images into vibrant RGB images. This grayscale-assisted approach significantly improves image quality for applications such as assisted driving and security surveillance.