Professor Ruimin Hu’team won the 21th Multimedia modeling international conference best student running up award.

After winning the lonely global best paper award of Pacific-Rim Conference on Multimedia in December last year , another good news arrived on January 4 to 7, 2015 at the 221th Multimedia modeling international conference ,held in Sydney, Australia .The team of Wuhan University computer school’s professor Ruimin Hu ‘s achievement “Coupled Discriminant Multi-Manifold Analysis (CDMMA) with Application to Low-Resolution Face Recognition” won the conference best student paper award. The paper is written by Wuhan University Computer School National multimedia software engineering center’s PHD student Jun jun Jiang, Professor Ruimin Hu ,lecturer Zhen Han, PHD Liang Chen and Professor Jun Chen together.

In recent years more than 300 billion yuan is invested in 660 cities in the country to carry out the construction of safe city video monitoring specific project .The world’s largest video surveillance network have been built, and it plays an increasingly important role in maintaining the safe of the city and striking illegal and criminal activities . In traditional criminal investigation process, the investigator always extract and analyze the surveillance video data following the principle of time and space .But the suspected target in city blocks monitor generally stay far from the surveillance cameras , and the human size in the video and resolution is very low. So it is difficult to achieve effective identification directly.

Low-Resolution Face Recognition(LRFR) is a real identity recognizing technology using the low quality and resolution face image. Although existing face recognition system under ideal conditions can achieve satisfactory recognition result , in the actual video surveillance applications the result of uncontrolled environment (such as pose ,illumination, expression change )and the object distance away from the surveillance cameras lead to the sharp drop of conventional face recognition methods’ identification effect.

With the support of national natural science foundation, 973 items, national science and technology support plan corpus, the ministry of public security key scientific research project, the team of Wuhan University computer school’s professor Ruimin Hu make a research on the problem of face recognition in low quality and resolution, focusing on analyzing sample of the manifold extracted from a single object in a different environment (different pose ,illumination, expression, etc.)and  constitution mechanism and regularities of distribution of the manifold sample between multiple objects. Set up the graph model in and between manifold. Projected the original high resolution face image sample onto a Discriminative public space, in which we make comparison and recognition. The proposed discriminate manifold analysis method increase around 15% on the international standard datasets compared with the traditional low spatial resolution of the identification method.

The team of Wuhan University computer school’s professor Ruimin Hu focus on the city security intelligent video analysis technology research for years, proposed and fully participated the implementation and construction of the national “safe city” major project and presided over the formulation of the first safe city technical standard. In recent years, in respect of the monitoring environment facing super-resolution and recognition professor Ruimin Hu has made significant achievements. National invention patent number ranks the first in the field of Face super-resolution. The “criminal video surveillance data analysis system” led by professor Ruimin Hu has been widely used in the police all over the country playing an important role in safeguarding national security and stability.

International Conference on Multimedia Modeling is the highly recommend conference for multimedia research field by china computer society, aimed at promoting the communication of up to date research and achievements in multimedia content analysis ,multimedia signal processing and communications and multimedia applications and services. The conference has received wide attention of the international academic since 1993.