From October 21 to 25, the 27th ACM International Conference on Multimedia (ACM MM) was held in Nice, France. The conference selected two of CVTE Research’s papers (1 full paper, and 1 challenge paper) and invited it to France to make an explanatory report.

As an annual event Initiated in 1993, ACM MM has become an international top multimedia conference and a Class A international academic meeting recommended by China Computer Federation. Its most researched subjects are multimode learning, large-scale image and video analysis, social media, multimode man-machine interaction, computational vision, computational image, etc.

Two of CVTE Research’s papers were selected by this year’s ACM MM. One of the papers researches the courseware originality measurement technique, and the other was the reverse image search technique for beauty products. The two papers are both from CVTE Research’s data mining group. On October 21, the data mining group sent five researcher representatives to the conference to make an explanatory report on the papers’ views and conduct academic discussions with scholars.  

Paper 1: “MOC: Measuring the Originality of Courseware in Online Education Systems”

Courseware originality measuring technique – The technique makes use of the multimode learning technique, the unsupervised mutual information maximization technique, and the integration of courseware’s images and textual information to intelligently mark courseware’s originality. By now, it has been applied to seewo’s education businesses to achieve large-scale automatic courseware review.

(The core views of the included paper)

Paper 2: “The retrieve of the beautiful: self-supervised salient object detection for beauty product retrieval”

Reverse image search technology for beauty products – The paper proposes a self-supervised salient object detection technique applied to learn the foreground representations of beauty product images to achieve the robust retrieval of beauty products, gathering the experience in research and practice of reverse image search.

(the core views of the included paper)

Actually, CVTE Research’s research achievements have been applied to actual businesses one after another, making each of the group’s business products (for example, smart interactive flat panel, smart display control, LCD TV mainboard, on-board display, and smart speaker) more competitive in the market.

In 2014, CVTE founded CVTE Research that gathers domestic and overseas experts and scholars and focuses on the researches of visual computing, speech signal information processing, data mining, natural language processing, medical signal analysis, robot perception and control, electric machines and drives, etc. CVTE Research has been qualified as a Postdoctoral Innovation Practice Base of Guangzhou and established extensive cooperation with domestic and overseas famous higher-educational institutions and research institutions.

(seven major research directions of CVTE Research)

It is a great honor to CVTE that CVTE Research’s papers were included by the ACM MM. In the future, CVTE will be determined to shoulder the mission of exploring the future, focus on scientific research innovation and professional development and continue maintaining its open and intercommunicated academic exploration attitude to endow CVTE Research with the enormous magnet for elite professionals and eternal vitality.