Multimodal Artificial Intelligence Research Center
Location:
The MUST LIU’s Innovation and Technology Center, ITC-14
Brief introduction：
This research center is established to address the key technologies in the next generation of artificial intelligence—specifically, multimodal perception learning and decision-making. The goal is to solve a series of critical scientific and application problems related to multimodal feature representation learning and the decision-making and planning based on it. Additionally, we aim to establish extensive connections with the industry to transfer our research outcomes into industrial applications.
Equipment:
The main product of this project is the appearance quality inspection machine and intelligent software system. The appearance quality inspection machine adopts the Huichuan robotic arm with two built-in 4K HD industrial cameras, which can be programmed to take multi-angle shots of the carrier object, and the self-developed intelligent software system can accurately and efficiently locate the target object, measure its dimensions and detect defects. The project provides a one-stop solution to address the current pain points of machine vision quality inspection, including high algorithm design costs, long lead time, low versatility and maintenance difficulties. A number of key technologies with independent intellectual property rights are integrated in the software system, including semi-automatic annotation technology for low manual labour, automatic model search technology supporting multi-indicator optimization, multi-source anomaly detection mechanism for open environment and sustainable model optimization strategy based on data drive. It has the advantages of high automation, low labour cost and strong detection performance compared with similar products, which can help customers improve product quality, increase production efficiency and reduce labour cost.
The main research projects include：
Research on Key Technologies for Digital Human Animation, 2024-2026.
Key Technology for General Visual Models,, 2021-2024.
Key Technology in Augmented Reality, 2020-2024.
Research and Development of Mobile Applications for Crowd Counting Based on Image Analysis, 2017-2019.
Intelligent Advertising Recommendation System: Key Technologies and Application Demonstration, 2017-2018.
Structured Analysis of Facial Features Based on Video Spatiotemporal Information Modeling, 2019-2022.
STEP Perpetual Learning based Collaborative Intelligence: Theory, Methodologies and Applications, 2020-2022.
Development and Industrialization of Core Engine Technology Platform for Virtual Reality, 2019-2023.
Multimodal Human Feature Recognition Algorithms and Systems, 2019-2022.
Application Research on Anomalous Crowd Behavior Analysis in Large-Scale Urban Spaces, 2019-2020.
Anti-Spoofing Detection Technology in Facial Recognition, 2018-2021.
