Team introduction

The intelligent image analysis team led by Professor Lu Huchuan of Dalian University of Technology has always adhered to the scientific research spirit of "seeking truth and innovation", and deeply integrated into the national and local development strategic needs. Over the years, it has been committed to the research in the field of computer vision and artificial intelligence, and has made a series of pioneering work in the direction of target tracking and significant detection.

Team promotion video

/system/go.jsp?treeid=1081&apptype=content&urltype=news.NewsContentUrl&param=wbnewsid%3D2053

Team leader

Professor Lu Huchuan: doctoral supervisor, winner of National Science Fund for Distinguished Young Scholars, young and middle-aged scientific and technological innovation leader of the Ministry of Science and Technology, and scientific and technological innovation leader of the Liaoning Talents Program (distinguished professor of Liaoning Province). Presided over a number of projects including major science and technology innovation projects of the Ministry of Science and Technology and key projects of the NSFC. Google cited more than 20000 times academically. It has won two second prizes of the Ministry of Education's Natural Science Award and several international academic awards.

Team members

Associate Professor Wang Dong	Wang Lijun Postdoctoral	Bo Chunjuan, Senior Engineer	Wang Yifan Lecturer

Team R&D direction

The team leader and members have been engaged in the research of computer vision, pattern recognition and image and video processing for nearly 20 years, and have made a series of influential scientific achievements in the fields of target tracking, saliency detection, image depth estimation, video and image segmentation, and abnormal behavior detection.

Team research achievements and application fields

The team published more than 200 high-level academic papers in the field of artificial intelligence and computer vision, cited more than 20000 papers, and won the CVPR2020 Best Paper Nomination Award, IET Best Paper Award and other awards.

In the past five years, the total contract amount of various scientific research projects approved is about 20 million yuan, and it has undertaken a number of scientific research projects such as the major scientific and technological innovation projects of the Ministry of Science and Technology of the People's Republic of China, the National Natural Science Foundation for Distinguished Young Scholars, joint funds, key funds, and overseas cooperation funds. Relevant research achievements won the second prize of Natural Science Award of the Ministry of Education in 2016 and 2019 respectively.

Cooperate with well-known enterprises at home and abroad on scientific research projects in many fields, such as facial expression recognition, multi-target tracking, long-term tracking, portrait segmentation, population density estimation, etc. Representative achievements are as follows:

1. Facial expression recognition: Cooperate with Konica Minolta in Japan to achieve fast and accurate facial expression recognition, facial key point detection and facial orientation judgment.

2. Shallow depth of field rendering: Cooperate with Adobe to achieve a rendering effect comparable to that of the SLR camera when shooting with an ordinary mobile phone.

3. Dance migration: cooperate with ByteDance American AI Lab to realize the migration of any person's actions to another person.

4. Portrait video segmentation: cooperate with ByteDance American AI Lab to achieve accurate segmentation of people in videos.

5. Abnormal behavior detection: cooperate with Omron to complete the abnormal behavior detection system under the monitoring scenario.

6. Multi pedestrian tracking: Cooperate with Omron Corporation of Japan to achieve multi person accurate tracking in multiple scenarios.

7. Video registration: Cooperate with Shenzhen Anjian Technology Co., Ltd. to develop multi CCD digital X-ray machine.

8. Embedded platform UAV pedestrian detection (third in the international competition)

9. Pedestrian tracking in low altitude UAV scene (building benchmark dataset and algorithm in this field)

Relevant algorithms have been applied to actual software products, and have been granted a number of invention patents, and won the VOT champion of the most authoritative international academic competition in the field of visual tracking from 2018 to 2020.

Partner