Research

Main Research Fields Introduction

Main Introduction shown in the team card.

Computer Vision
Medical Image Analysis
Machine Learning
Industrial Internet Security

1. Computer Vision [Detail]

How to observe and understand the world is one essential task in computer vision. Our team works on finding solutions to 3D visual understanding, especially on 3D object recognition and retrieval. Current projects include multi-view object recognition for self-driving vehicle, 3D multi-modal (point cloud, multi-view, volumetric and mesh) data fusion, high-speed visual reconstruction and visual detection.

2. Medical Image Analysis [Detail]

This part of work mainly focuses on medical image processing and computer-aided diagnosis. In medical image processing, we design effective and efficient deep learning methods for brain image segmentation and registration. Our research designs new methods for diagnosis of brain degenerated diseases, such as Alzheimer’s Disease, and cardiovascular disease, supported by National Key R&D Program of China.

3. Machine Learning [Detail]

Complex data correlation modelling and representation plays an important role in many applications, such as social media analysis, data classification and medical diagnosis. Here, we focus on graph/hypergraph based learning, hypergraph neural networks, multi-modal data fusion, metric learning and cost-sensitive learning methods, and their applications on visual classification, software defect prediction, and social media data filtering and recommendation.

4. Industrial Internet Security

With the development of Industrial Internet and Industrial 4.0, industrial control area attaches more and more attentions from researchers. Nowadays, big data, cloud computing and Internet of things have been the kernel techniques for the national critical infrastructures. In this part, we focus on complex relationship description, anomaly detection and model adaptation.