See the virtual world like the real world? Development of image deep learning technology that presents VR and AR screens more vividly and realistically

- DGIST Professor Jin Kyong-hwan’s research team develops algorithm that has 3dB higher resolution than existing algorithms and compatible with various warping domains - Published in ‘ECCV’, an academic journal with global authority in the field of vision technology

□ Professor Jin Kyong-hwan's research team of the Department of Electrical Engineering and Computer Science, DGIST (President: Kuk Yang) developed image processing deep learning technology that reduces memory speed and increases resolution by 3dB compared to existing technologies. Developed through the joint research with Master Choi Kwang-pyo of Samsung Research, this technology reduces aliasing phenomenon on the screen compared to the existing “signal processing-based image interpolation technology (Bicubic interpolation),” thus producing more natural video output. In particular, it can restore the high-frequency part of images clearly. It is expected to display a natural screen when using VR or AR.


□ The “signal processing-based image interpolation technology (Bicubic interpolation)” preserves desired images in various environments by designating a specific location of an image. It has the advantage of saving memory and speed, but deteriorates the quality and deforms the image.


□ To address this issue, deep learning-based ultra-high-resolution video image conversion technologies were developed, but most of them are convolutional artificial intelligence network-based technologies, which have the disadvantage of inaccurate estimation of values ​​between pixels, which can lead to image deformation. The implicit expression neural network technology to overcome these disadvantages is drawing attention, but the implicit expression neural network technology has the disadvantage that it cannot capture high-frequency components and increases memory and speed.


□ DGIST Professor Jin Kyong-hwan's research team developed a technology that resolves the image into several frequencies so that the characteristics of high-frequency components can be expressed in the image and reassigns coordinates to resolved frequencies using the implicit expression neural network technology so that the image can be shown more clearly. It can be described as a new technology that combines the Fourier analysis, which is an image deep learning technology, and implicit expression neural network technology. The newly implemented technology can make improvements to the disadvantage of the implicit expression neural network that could not restore high-frequency components by resolving essential frequency components in restoring images through an artificial intelligence network.


□ Professor Jin Kyong-hwan said, “The technology developed this time is excellent as it shows higher restoration performance and consumes less memory than technology used in the existing image warping field. We hope that the technology is utilized in for image quality restoration and image editing in the future and hope that it will contribute to both academia and industries.”


□ Meanwhile, this research achievement was made possible through the support of the National Research Foundation of Korea (NRF), the Institute for Information & Communication Technology Planning & Evaluation (IITP), and DGIST. It was published in the world-renowned academic journal 'ECCV (European Conference on Computer Vision)' in the field of vision technology.

 Correspondent author's email address : [email protected]

Published: 14 Dec 2022

Contact details:


333, Techno jungang-daero, Hyeonpung-myeon, Dalseong-gun, Daegu, 42988

News topics: 
Academic discipline: 
Content type: