Research

Journal recommendation! Professor Li Jianqiang's team: Real-time comprehensive defense strategy for neural networks based on compression perception

Professor Li Jianqiang's team and collaborators published a paper titled "Real-time Comprehensive Defense Strategy Based on Compressed Sensing Neural Networks" in *Chinese Journal of Computers*, a leading academic journal in the field of computer science and technology and a CCF-recommended Class A Chinese scientific and technological journal. The paper was also featured twice on the *Chinese Journal of Computers* WeChat official account: 1) AI Security | Real-time Comprehensive Defense Strategy Based on Compressed Sensing Neural Networks (qq.com); 2) Recommended Article: Real-time Comprehensive Defense Strategy Based on Compressed Sensing Neural Networks (qq.com) (Click the link to view). This work points out that in recent years, visual recognition models based on deep neural networks have gained popularity due to their advantages in accuracy, cost, and...

Key technologies for facial recognition analysis

Key technologies for face recognition and analysis have received numerous accolades, including the Guangdong Provincial and Shenzhen Municipal Natural Science Awards, the Wu Wenjun Artificial Intelligence Natural Science Award, and the Chinese Institute of Electronics Natural Science Award. The key technologies used in this research include:

Basic theory of ubiquitous perception system and industrial application

The fundamental theory of ubiquitous sensing systems includes intelligent sensing, intelligent wireless communication, heterogeneous interconnection, and massive multi-mode IoT sensing data processing. Combining fifth-generation communication technology and artificial intelligence, it endows everything with sensing capabilities through the network, ultimately generating massive amounts of data in the cloud and enabling intelligent applications. This achievement has won awards including the First Prize of Guangdong Provincial Science and Technology Award, the Second Prize of Natural Science Award from the Ministry of Education, and the Second Prize of Guangdong Provincial Science and Technology Progress Award.

Virtual human driving and synthesis technology

The Visual Big Data Research Center studies virtual human synthesis technology, synthesizing realistic virtual human videos based on spoken text and speech, and combining this with large language models to synthesize virtual humans that can automatically answer questions. Building upon a series of face generation technologies, the center also researches real-time driving technology for 3D human bodies/faces; and develops interactive facial responses that understand the other person's emotions and generate corresponding facial expressions.

Face recognition/analysis and generation

The Visual Big Data Research Center studies facial recognition algorithms, algorithms for discriminating various facial attributes such as expressions, depression, and personality, face-swapping algorithms based on generative adversarial networks and diffusion models, and face generation algorithms based on text descriptions. Their facial recognition algorithm won second place in the ICPR accuracy test. Their latest 200-layer neural network facial recognition algorithm achieved industry-leading accuracy of 99.27% ​​and 99.19% in the MegaFace million-test 1:N and 1:1 ratios, respectively. Its overall performance exceeded 98% in the MFR Ongoing Challenge for masked faces, also currently the highest in the industry. Their facial motion unit detection algorithm won the 5th Annual Meeting of the CVPR 2023...

AI-based cloud-based multidimensional health data intelligence platform

my country's population development is facing a turning point, with a significant slowdown in total population growth, a continued decline in the birth rate, and a deepening aging population, making health and social issues related to the elderly and children increasingly prominent. Approximately 900,000 new cases of birth defects are reported annually in my country, significantly impacting the quality of the newborn population and maternal and child health; over 78% of the elderly suffer from at least one chronic disease, and the number of disabled elderly will continue to increase, causing incalculable socioeconomic losses. This project, based on technologies such as artificial intelligence, the Internet of Things, and big data, constructs a cloud-plus-device multi-dimensional health data intelligent platform, utilizing portable intelligent medical terminals, mobile applications, and...

Big Data Visualization and Analysis Decision System

The big data visualization and analysis decision-making system encompasses a variety of functions, including large-scale data acquisition and optimization fusion, multi-scale geometric modeling and image processing, visual content generation and simulation rendering, complex scene reconstruction and recognition understanding, 3D collaborative perception and human-computer interaction, simulation learning and enhanced cognition, and massive information visualization and visual analysis. This achievement is based on the computer graphics program at Shenzhen University, which ranks second in China and third in Asia.

IoT-based health big data application development platform

The laboratory has established an IoT-based health big data application platform based on the integration of big data from clinical practice, IoT terminals, and public health. This platform enables the analysis and prediction of chronic disease and maternal and infant health data. This project is a National Key Research and Development Program project, and the technologies used are mainly multi-source data collaborative sensing, including: ① an adaptive composite filter based on hardware and software collaborative filtering; ② a biometric point extraction method based on dynamic thresholds; ③ simplified calculations, suppression of dynamic noise interference, and ensuring data accuracy. Related results have been published in journals and conferences such as IEEE TII and IOTJ. This achievement won the national Maker Competition championship and...

Gene big data compression storage and standardization

In the project on gene big data compression, storage, and standardization, the laboratory led the establishment of the national independent standard technical system for genome sequencing data compression (AVS-G), established a standard test database for gene sequencing data compression, and developed reference software for gene sequencing data compression.

RSP Big Data Analytics Platform

The RSP big data analytics platform adopts the approach of "breaking down large data into smaller, smaller data representing large data, and analyzing large data from small data," transforming massive datasets into independent, identically distributed RSP data blocks. This solves the problem of the incompatibility between data analysis efficiency and accuracy.