Xiaojuan Li

Dr. Xiaojuan Li

orcid

Research Focus & Metrics

Academic Career Length

18Years (2009-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 March 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
YWYonggu Wang
Yonggu Wang
KPKailin Pan
Kailin Pan
YSYifan Shao
Yifan Shao
JMJiarong Ma
Jiarong Ma
XLXiaojuan Li
Xiaojuan Li
With advances in digital technology, including deep learning and big data analytics, new methods have been developed for autism diagnosis and intervention. Emotion recognition and the detection of autism in children are prominent subjects in autism research. Typically using single-modal data to analyze the emotional states of children with autism, previous research has found that the accuracy of recognition algorithms must be improved. Our study creates datasets on the facial and speech emotions of children with autism in their natural states. A convolutional vision transformer-based emotion recognition model is constructed for the two distinct datasets. The findings indicate that the model achieves accuracies of 79.12% and 83.47% for facial expression recognition and Mel spectrogram recognition, respectively. Consequently, we propose a multimodal data fusion strategy for emotion recognition and construct a feature fusion model based on an attention mechanism, which attains a recognition accuracy of 90.73%. Ultimately, by using gradient-weighted class activation mapping, a prediction heat map is produced to visualize facial expressions and speech features under four emotional states. This study offers a technical direction for the use of intelligent perception technology in the realm of special education and enriches the theory of emotional intelligence perception of children with autism.

1 December 2009 in Progress in Natural Science: Materials International
Verified
Journal Article
Open access
DMDan Meng
Dan Meng
XLXiaojuan Li
Xiaojuan Li
WZWenji Zhao
Wenji Zhao
HGHuili Gong
Huili Gong

20 February 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
YWYonggu Wang
Yonggu Wang
KPKailin Pan
Kailin Pan
YSYifan Shao
Yifan Shao
JMJiarong Ma
Jiarong Ma
XLXiaojuan Li
Xiaojuan Li
With advances in digital technology, including deep learning and big data analytics, new methods have been developed for autism diagnosis and intervention. Emotion recognition and the detection of autism in children are prominent subjects in autism research. Typically using single-modal data to analyze the emotional states of children with autism, previous research has found that the accuracy of recognition algorithms must be improved. Our study creates datasets on the facial and speech emotions of children with autism in their natural states. A convolutional vision transformer-based emotion recognition model is constructed for the two distinct datasets. The findings indicate that the model achieves accuracies of 79.12% and 83.47% for facial expression recognition and Mel spectrogram recognition, respectively. Consequently, we propose a multimodal data fusion strategy for emotion recognition and construct a feature fusion model based on an attention mechanism, which attains a recognition accuracy of 90.73%. Ultimately, by using gradient-weighted class activation mapping, a prediction heat map is produced to visualize facial expressions and speech features under four emotional states. This study offers technical direction for the use of intelligent perception technology in the realm of special education and enriches the theory of emotional intelligence perception of children with autism.

1 December 2020
Verified
Conference Paper
Closed access
RWRui Wang
Rui Wang
YGYong Guan
Yong Guan
XLXiaojuan Li
Xiaojuan Li
RZRui Zhang
Rui Zhang
Cyber physical system (CPS) is a multi-dimensional complicated system integrating computing, communication and physical environment. CPS is widely used in safety-critical areas such as aerospace, intelligent transportation and medical equipment. So ensuring the security and reliability of CPS is of great significance. Formal verification is one of the useful ways. This paper builds timed automata models for the communication process of CAN bus used in CPS. Our research especially analyses the gateway in the communication process, and simulates the transmission with different rates between the external environment and internal unit. The task also takes into account the packet transmission priority. The model checking tool Uppaal is used to verify the functional and real-time properties. The verification results illustrate that the established model can meet the relevant properties, and the packet can be transmitted in an orderly and efficient manner.

1 May 2009
Verified
Conference Paper
Closed access
ZGZhaoning Gong
Zhaoning Gong
HGHuili Gong
Huili Gong
XLXiaojuan Li
Xiaojuan Li
WZWenji Zhao
Wenji Zhao
YZYou-Quan Zhang
You-Quan Zhang
By utilizing TM/ETM+ remote sensing images from 1984 to 2004 as the major data sources, combining with the field survey to get the rule of dynamic process of Beijing wetland ecosystem, this paper systematic analyzed the hydrology, soil, plant condition of the typical wetland area- Widgeon Lake Wetland Natural Reserve in Beijing. And the results are as follows: In recent years, wetland areas were sharply reduced by half. The negative ecological effects around wetland areas have a close relationship with the deterioration of wetland resources. Salinized area had increased every year and reached to a head in 2004, which accounted for about 21.79 square kilometers. The serious and medium level salinity areas were in the majority in region. Due to the soil salinization, the marsh vegetation species changed to the mesophytic, xeromorphic and salt tolerable ones. Along with the continuously shrinking of water area, wetland dominated the low coverage. The area of medium to high and high coverage in 2004 reduced 71.09 percent and 99.26 percent respectively compared with those in 1998. The regional vegetation cover had shown the degradation trend. Accordingly the soil conservation function of vegetation weakened and soil fertility declined. The cold-humid effect of wetland had been impaired for the breakage of wetland environment in recent years, which led to the drying and warming tendency of regional microclimate. The phenomenon of the decrease amount of regional daily evapotranspiration proved that the climate was drying and warming. The serious disequilibrium between source and sink landscape in every sub-watershed of wetland was an important factor to cause non-point source pollution. The nonpoint source pollution in upper reaches of wetland had caused the high eutrophic water in recent years. From the field water quality monitor we could find the main contaminations were total nitrogen (TN), total phosphorus (TP), chemical oxygen demand (COD) and fluorid (F). Water quality has reduced up to IV or V grade. The wetland landscape shape fragmentation indexes were all over 0.99 from 1998 to 2002, which showed the impacts of human disturbance on landscape was intensive, the wetland landscape had suffered a serious destruction and the ecology was instability. Based on the above analysis, the regional ecological negative effects were closely connected with the decrease of wetland resources. We should bring forward corresponding effective protection and utilization measures and give the scientific basis for the restoration of wetland resources.

26 September 2009
Verified
Conference Paper
Closed access
PDF file
CYCankun Yang
Cankun Yang
XLXiaojuan Li
Xiaojuan Li
QLQiang Liu
Qiang Liu
HZHuimin Zhao
Huimin Zhao
JZJia Zhang
Jia Zhang
HZHaibo Zhang
Haibo Zhang
This paper aims to introduce a case of Socio-economic statistical Spatio-temporal Database. This database system services in the rural socio-economic statistical work, which is a combination of statistical tables, spatial data, search algorithm and maintenance interface. Administrative codes are the conjunction media of spatial data and attribute data, and also are the key words of database query processing. Through storing the changing information in the database, it could reflect the change of administrative divisions. As the main issues of database design, the studying of the approach to recording and querying these changes as well as the processing of statistical data by the rules of administrative divisions changes, requires a large amount of research work. To address these problems, a series of management analysis tools have been developed to deal with the processing of socio-economic statistical data with changes in the administrative division. A searching algorithm of spatio-temporal database is used to ensure the comparability of the results, which are acquired by the positive sequence and the anti-sequence temporal query under complex spatial changes in the administrative division. According to the spatial changes, searching algorithm of spatio-temporal database mainly translates temporal series statistical data into standard format data which is matched to the benchmark year. The searching algorithm controls the process of inquiry through recursion of the table of the administrative code changes, which are composed of multi-way tree structure and double linked list and record the relationship between upper and lower level administrative units. These search algorithms and meta-data storage structures constitute the spatio-temporal database, so as to serve the spatial analysis of statistical data. The comparability problem mentioned above was well solved by this approach. And a set of functions was provided by this system with spatio-temporal database, such as specialization of statistical data, temporal query, spatial data which can be automatically updated, and maintenance interface.