The Human and Vision Lab, led by Professor Ryu Han-chul of the Department of Artificial Intelligence Convergence at Sahmyook University, has successfully published a workshop thesis at the European Computer Vision Conference (ECCV) 2022. ECCV is considered one of the world’s most prestigious artificial intelligence (AI) conferences, along with Computer Vision and Pattern Recognition Conference (CVPR) and International Conference on Computer Vision (ICCV).
The thesis title is CAIR: Fast and Lightweight Multi-Scale Color Attention Network for Instagram Filter Removal. Yeo Un-ha, a Ph.D. research fellow, undergraduates Oh Wang-taek and Kim Young-il, and Dr. Kang Kyung-soo of the Construction Technology and Management Research Institute (Director Ryu Han-gook, Professor of Architecture) participated as well. The research team proposed a method to effectively remove the filter added to the image based on the experience of performing the Terahertz image noise removal research task.
The artificial intelligence network proposed by the research team showed excellent performance in restoring the original color by focusing on color information and having a faster inference speed than other algorithms by reducing the structure. This technology is expected to be highly utilized in the image preprocessing process to improve the performance of various computer vision tasks, such as restoration of degraded images, image resolution improvement, object detection, and image classification.
Professor Ryu’s Human & Vision Lab is a new artificial intelligence research team composed of one doctoral fellow and seven undergraduate students. They carry out tasks of the Korea Research Foundation and the Korea Electronics and Telecommunications Research Institute, conduct artificial intelligence-based construction safety and digital healthcare convergence research to derive competitive research results, and cultivate excellent artificial convergence human resources.
- There are no registered attachments.