Research Direction

1.Big Data Intelligence Algorithms (Trustworthy AI and AI for Sciences)

2.Edge Intelligence Technology and Its Applications

3.Intelligent Robotics Technology and Systems

Team

The center currently has 15 researchers, including 4 senior researchers, 7 associate senior researchers, 3 intermediate researchers, and 1 postdoctoral researcher. There are also 80 graduate students currently enrolled.

Center Director: Shang Mingsheng (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202012/t20201225_5842652.html)

Deputy Center Director: Zhang Xuerui (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202012/t20201225_5842640.html)

Key Academic Members:

Lin Yuanchang (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202101/t20210127_5878264.html)

Shi Xiaoyu (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202012/t20201225_5842672.html)

Shan Kun (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202101/t20210127_5878270.html)

Chen Lin (Hyperlink: http://cigit.cas.cn/sourcedb/rcdw/202303/t20230313_6696802.html)

Zhou Botian (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202202/t20220216_6355972.html)

Zhang Fan (Hyperlink: http://sourcedb.cigit.cas.cn/rcdw/202012/t20201225_5842638.html)

Yao Yuan

Representative Research Achievement

The research achievements of the Center have been successfully applied in various industries and fields. It has nurtured three talents at the national level and fourteen talents at the provincial and ministerial levels. The Center has led four major national and corporate projects with a funding of over tens of millions of yuan, and has been involved in the formulation of eleven national and local standards. It has obtained more than 130 authorized invention patents, published three monographs, and contributed over 100 academic papers to prestigious journals such as CAS first-tier journals or CCF A/B class journals. The Center has been awarded four first prizes, including the Chongqing Natural Science Award, Science and Technology Progress Award, and the China Association for Artificial Intelligence Award.

As the supporting unit, the Center has established the Online Monitoring Center for the Ecological and Environmental Monitoring System of the Three Gorges Project under the State Council's Three Gorges Office. It has also established the Chongqing Key Laboratory of Edge Intelligence Computing and actively participated in the construction of several national and provincial-level scientific research platforms, including the National Key Laboratory for Mountain Disasters and Engineering Safety, and the Chongqing Key Laboratory for Common Technologies and Systems of Service Robots.

(A) Representative Achievements

1. Big Data Intelligent Computing Theory and Methods

In response to the challenges faced by intelligent decision-making in uncertain environments, such as spurious correlations bias interference, out-of-distribution (O.O.D) generalization, and decision fairness, the center has made significant contributions. Taking interactive recommendation decision-making as an example, the specific impact mechanisms of exposure bias on interactive decision-making have been analyzed. The mathematical mechanisms underlying assimilation and contrast psychological effects in user rating behavior have been revealed. A dual-perception interactive recommendation approach that considers diversity and novelty has been proposed. A guided recommendation system focused on long-term fairness has been established, enabling robust decision-making in large-scale uncertain environments. These advancements effectively alleviate the issues of information filter bubbles and internet addiction caused by recommendation systems and provide a fresh perspective for studying fairness in AI decision-making.

2. Big Data Platform Technology and Applications

The center has made breakthroughs in several key technologies related to big data, including the integration of massive heterogeneous data, efficient management of high-performance data center resources, and lightweight edge intelligence. Techniques such as multidimensional cropping based on spatiotemporal data and video motion expression analysis have been proposed. These techniques enable intelligent recognition and behavior analysis of "fast-moving smaller targets" in "larger scenes," addressing the challenges of real-time motion quantification analysis in sports events. They have been successfully applied in the 2022 Beijing Winter Olympics. Comprehensive recognition techniques for typical targets in complex environments, considering multiple perception modalities, have been developed. These techniques facilitate detection, recognition, tracking, analysis, and retrieval of personnel and vehicles in extremely harsh and complex environments for security applications. They also enable the monitoring and discrimination of group behavior and specific behaviors. The developed "edge-edge-cloud" big data platform enables effective integration, adaptation, and rapid flexible deployment of various AI models on domestic heterogeneous chips.

3. Fundamental Theories, Evaluation System, and System Applications of Intelligent Robots

Significant advancements have been made in robot control technology based on unknown structures, enabling high-precision control of robots. A unified framework for multi-robot competition and cooperation has been proposed, achieving collaborative algorithms for multi-robot systems under limited communication. A robot motion planning framework based on neural dynamics has been developed, enabling efficient control of robots. Testing methods and error compensation mechanisms for angle, torque, temperature, and vibration in gear detectors have been proposed. Comprehensive performance evaluation systems for gears and quality control systems for production processes have been established, forming a comprehensive performance assessment model for robots. An intelligent operational robot system has been developed, integrating flexible manipulators, force sensing, machine vision, and motion control. This system has been applied in automated production lines for flexible and precise assembly, intelligent welding of non-standard medium-thick plates, and complex surface polishing of irregular workpieces. An all-terrain electric track unmanned agricultural machinery power platform, covering the entire process of orchard operations, has been developed. This platform realizes a fully automated intelligent system from agricultural situation discovery, diagnosis, and decision-making to operation execution.

4. Automatic Monitoring, Intelligent Analysis, and Platform Visualization of Key Elements in Aquatic Ecological Environments

Breakthroughs have been made in detection technologies for key elements in aquatic ecological environments, data-driven analysis technologies that integrate domain knowledge, and the construction of a water ecological environment monitoring and early warning cloud platform. Conventional hydrological and water environmental automatic monitoring devices have been improved, and in-situ monitoring devices for core indicators in aquatic ecosystems have been innovatively developed. A water ecological environment monitoring and early warning information cloud platform has been established. These advancements have been applied in various aspects such as water environment monitoring in the Three Gorges Reservoir area, monitoring of the YueLai Sponge City, flood forecasting in Qijiang, and water ecological perception in Dianchi Lake. The research has achieved the integration and innovation of the entire technical chain in the development and application of water ecological environment automatic monitoring devices and systems.

(B) Representative Papers

[1] Xiaoyu Shi, Quanliang Liu, Hong Xie, Di Wu, Bo Peng, Mingsheng Shang, Defu Lian. Relieving Popularity Bias in Interactive Recommendation: A Diversity-Novelty-Aware Reinforcement Learning Approach. ACM Trans. Inf. Syst. 42(2): 52:1-52:30 (2024)

[2] Hong Xie, Mingze Zhong, Xiaoyu Shi, Xiaoying Zhang, Jiang Zhong, Mingsheng Shang. Probabilistic Modeling of Assimilate-Contrast Effects in Online Rating Systems. IEEE Trans. Knowl. Data Eng. 36(2): 795-808 (2024)

[3] Mei Liu, Mingsheng Shang. Orientation Tracking Incorporated Multicriteria Control for Redundant Manipulators With Dynamic Neural Network. IEEE Trans. Ind. Electron. 71(4): 3801-3810 (2024)

[4] Jiawang Tan, Mingsheng Shang, Long Jin. Metaheuristic-Based RNN for Manipulability Optimization of Redundant Manipulators. IEEE Trans. Ind. Informatics 20(4): 6489-6498 (2024)

[5] Di Wu, Bo Sun, Mingsheng Shang. Hyperparameter Learning for Deep Learning-Based Recommender Systems. IEEE Transactions on Services Computing, 16(4): 2699-2712 (2023)

[6] Mei Liu, Liangming Chen, Xiaohao Du, Long Jin, Mingsheng Shang. Activated Gradients for Deep Neural Networks. IEEE Trans. Neural Networks Learn. Syst. 34(4):2156-2168(2023)

[7] Ye Yuan, Xin Luo, Mingsheng Shang, Zidong Wang. A Kalman-Filter-Incorporated Latent Factor Analysis Model for Temporally Dynamic Sparse Data. IEEE Trans. Cybern. 53(9): 5788-5801 (2023)

[8] Kun Shan, Ouyang Tina, Xiaoxiao Wang, Hong Yang, Botian Zhou, Zhongxing Wu and Mingsheng Shang. Temporal prediction of algal parameters in Three Gorges Reservoir based on highly time-resolved monitoring and long short-term memory network. Journal of Hydrology. (2022).

[9] Di Wu, Xin Luo, Mingsheng Shang, Yi He, Guoyin Wang, Xindong Wu. A Data-Characteristic-Aware Latent Factor Model for Web Services QoS Prediction. IEEE Trans. Knowl. Data Eng. 34(6): 2525-2538 (2022)

[10] Long Jin, Yimeng Qi, Xin Luo, Shuai Li, Mingsheng Shang. Distributed Competition of Multi-Robot Coordination Under Variable and Switching Topologies. IEEE Trans Autom. Sci. Eng. 19(4): 3575-3586 (2022)

(C) Representative Projects

[1] Intelligent Terminal Software, Collaborative Research Project by the Ministry of Industry and Information Technology, July 1, 2020, to July 31, 2023, Project Budget: 87 million RMB.

[2] YueLai New Town Sponge City Monitoring and Information Platform, March 14, 2017, to December 30, 2020, Project Budget: 37 million RMB.

[3] Three Gorges Reservoir Area Water Ecological Environment Perception System and Platform Operation, January 1, 2014, to June 30, 2019, Project Budget: 17 million RMB.

[4] Chongqing Yubei District Comprehensive Governance Information Platform, January 8, 2019, to September 30, Project Budget: 17 million RMB.

[5] Comprehensive Multi-Role Intelligent Interaction and Training System, National Key Research and Development Program, July 1, 2020, to July 31, 2022, Project Budget: 16 million RMB.

[6] Developing and applying precision variable-rate fertilization robots for mountain citrus, March 21, 2023-03 to November 1, 2025, 5 million RMB.

[7] Intelligent Collaborative Technology Support System and Application Demonstration for Political and Legal Affairs, National Key Research and Development Program, October 1, 2020, to December 1, 2023, Project Budget: 2.2 million RMB.

[8] Research and application of software-defined intelligent terminal systems, Chongqing Municipal Science and Technology Plan Project, January 1, 2022, to December 31, 2023, Project Budget: 2 million RMB.

[9] Research and application of key technologies for multimodal generation of large-scale models, Chongqing Municipal Science and Technology Plan Project, December 10, 2023, to November 30,2026, Project Budget: 2 million RMB.

[10] Development and Application of High-Precision Comprehensive Performance Testing Instrument for Precision Gear Reducers, National Key Research and Development Program, July 1, 2017, to June 30, 2021, Project Budget: 1.9 million RMB.

Contact Us 

Mr. Gan

Email: ganjie@cigit.ac.cn