National Center for High Performance Computing, NARLabs
National Center for High-performance Computing (NCHC), founded in 1991, is Taiwan's only national-level supercomputing center. The NCHC plays a leading role in cloud technology that possesses a large computing, storage, networking and platform integration, providing the domestic users with the Could Integration Service of High Performance Computing (HPC), high quality networking, high efficiency storage, big data analysis and scientific engineering simulation.
The goal of the NCHC is to become an internationally renowned HPC center that promotes scientific discovery and technological innovation. Since its inception, the NCHC has been dedicated to strengthening Taiwan's HPC and networking infrastructure. The NCHC has planned and implemented pilot research programs in HPC, cloud computing, as well as big data processing methods and applications. The NCHC provides professional technologies and platform services to academia, government, and industry, and helps to cultivate domestic talent in HPC-related fields.
In order to effectively support Taiwan's technology research, the NCHC constructed technology R&D platforms to support domestic and foreign R&D teams in developing HPC and big data applications, which cover engineering and science, environmental and disaster prevention, biomedicine, and digital cultural content creation, aiming to become a first-rate High-performance Computing Center.
AI Video Analysis and Retrieval System
Solution Description
Video surveillance systems are deployed at many public spaces for Urban security and surveillance purposes such as airports, train stations, and shopping malls. However, it is laborious to analysis and retrieval for specific persons in multiple camera surveillance systems, especially in the cluttered background and appearance variations among multiple cameras. In order to address the problem, this system presents an AI video analysis and retrieval method to extract the human attribute via deep-learning instance-segmentation technology. It uses attributes like clothes color and type to describe a person. The proposed system of person retrieval consists of four steps: (i) using deep-learning instance-segmentation technology to perform pixel-wise person segmentation; (ii) appearance-based attribute features with multi-CNN; (iii) search engine with fundamental attributes matching approach; and (iv) the video summarization technique to produce temporal abstraction of retrieval objects.