作者机构:
[Chang, CH; Ho, Chia-An; Chang, Chun-Hao; Yeh, Hung-Chih; Chen, Chao-Yuan; Ho, Chin-Shan] Natl Taiwan Sport Univ, Grad Inst Sports Sci, Taoyuan City, Taiwan.;[Li, Fang] Cent China Normal Univ, Sch Phys Educ, Wuhan, Hubei, Peoples R China.;[Chen, Chao-Yuan] Natl Taipei Univ Business, Sport Off, Taipei City, Taiwan.
通讯机构:
[Chang, CH ] N;Natl Taiwan Sport Univ, Grad Inst Sports Sci, Taoyuan City, Taiwan.
关键词:
Mechanical property;Myometer;Rowing;Symmetry index
摘要:
BACKGROUND: The mechanical properties of muscles, such as changes in muscle tone and stiffness, are related to sports performance and injuries. Rowers are at increased risk of muscle fatigue and injury during high-repetition and heavy-load cyclic muscle actions. In view of this, the aim of the present study was to investigate the acute effect on muscle tone and stiffness, as well as bilateral muscle asymmetry, in high school rowers after a 2000-meter rowing ergometer test. METHODS: Twelve young male rowers (age = 17.1 ± 0.9 years, body weight = 73.5 ± 9.7 kg) were included in the study. The data of muscle tone (frequency) and stiffness of the posterior deltoids (PD), latissimus dorsi (LD), and rectus femoris (RF) (dominant and non-dominant side) before and after a 2000-m rowing ergometer test were collected using a handheld MyotonPRO device. RESULTS: After the rowing ergometer test, the muscle tone of dominant side PD, LD, and RF were significantly increased (p < 0.05). On the other hand, the muscle stiffness of the non-dominant side LD and RF, as well as the dominant side PD, LD, and RF were significantly increased after the rowing ergometer test (p < 0.05). The muscle tone and stiffness results showed that the dominant side PD, LD, and RF were all significantly higher than the non-dominant side after the rowing ergometer test (p < 0.05), where bilateral PD and RF exhibits moderate asymmetry (5% < symmetry index < 10%). CONCLUSIONS: After a high-intensity and high-load 2000-m rowing ergometer test, PD, LD, and RF showed increases in muscle tone and stiffness, as well as changes in the symmetry of bilateral muscle mechanical properties.
期刊:
Journal of Organizational and End User Computing,2024年36(1) ISSN:1546-2234
通讯作者:
Peng, Y
作者机构:
[Xiong, Li; Chen, Yuanyuan] Cent China Normal Univ, Wuhan, Peoples R China.;[Peng, Y; Xiong, Li; Peng, Yi] Yangtze Univ, Jingzhou, Peoples R China.;[Ghadi, Yazeed Yasin] Al Ain Univ, Al Ain, U Arab Emirates.
通讯机构:
[Peng, Y ] Y;Yangtze Univ, Jingzhou, Peoples R China.
摘要:
This study aims to enhance the efficacy of personalized learning paths by amalgamating transformer models, generative adversarial networks (GANs), and reinforcement learning techniques. To refine personalized learning trajectories, the authors integrated the transformer model for enhanced information assimilation and learning path planning. Through generative adversarial networks, the authors simulated the fusion and interaction of multi-modal information, refining the training of virtual teaching assistants. Lastly, reinforcement learning was employed to optimize the interaction strategies of these assistants, aligning them better with student needs. In the experimental phase, the authors benchmarked their approach against six state-of-the-art models to assess its effectiveness. The experimental outcomes highlight significant enhancements achieved by the authors' virtual teaching assistant compared to traditional methods. Precision improved to 95% and recall to 96%, and an F1 score exceeding 95% was attained.
期刊:
Journal of Plant Research,2024年:1-13 ISSN:0918-9440
通讯作者:
Qin-Zheng Hou
作者机构:
[Nurbiye Ehmet; Xiang Zhao; Tai-Hong Wang; Yi-Ping Zhang; Kun Sun] College of Life Sciences, Northwest Normal University, Lanzhou, 730070, Gansu, China;[Nurbiye Ehmet] College of Life Sciences, Central China Normal University, Wuhan, 430079, Hubei, China;[Qin-Zheng Hou] College of Life Sciences, Northwest Normal University, Lanzhou, 730070, Gansu, China. hou_qzh@nwnu.edu.cn
通讯机构:
[Hou, Qin-Zheng] C;College of Life Sciences, Northwest Normal University, Lanzhou, 730070, Gansu, China.
摘要:
The intervention of nectar robbers in plant pollination systems will cause some pollinators to modify their foraging behavior to act as secondary robbers, consequently adopting a mixed foraging strategy. The influence of nectar robbing on pollinator behavior may be affected by spatio-temporal difference of robbing intensity, and consequently, may have different effects on the pollination of host plants. However, whether and how the nectar robbing might influence pollinators under different robbing intensity still needs further investigation. In this study, Symphytum officinale was used to detect the effect of nectar robbers on pollinators under different robbing intensity as well as their effects on plant reproductive success. Six robbing levels and three bumblebees with mixed foraging behaviors were used to evaluate the effect of different robbing intensity on pollinator behavior, visitation rate, flower longevity and pollen deposition. Our results indicated that the robbing rate increased gradually with the proportion of robbed flowers, but which did not affect the frequency of legitimate visits. The increase of robbing rate promoted the corolla abscission, and then enhanced the self-pollen deposition, but which had no significant effect on cross-pollen deposition. These results indicate that the overall fitness of S. officinale was improved by combined self and cross-pollination modes when visited by both pollinators and nectar robbers simultaneously. Although nectar robbing is not uncommon, its consequences for pollination in the interaction web have not been well studied. Our results emphasize the significance of indirect impacts in mediating the adaptive outcomes of species interactions.
期刊:
Journal of Inverse and Ill-posed Problems,2024年 ISSN:0928-0219
通讯作者:
Xiang, JL
作者机构:
[Xiang, Jianli; Xiang, JL] China Three Gorges Univ, Coll Sci, Three Gorges Math Res Ctr, Yichang 443002, Peoples R China.;[Yan, Guozheng] Cent China Normal Univ, Sch Math & Stat, Wuhan 430079, Peoples R China.;[Yan, Guozheng] Hubei Key Lab Math Phys, Wuhan, Peoples R China.
通讯机构:
[Xiang, JL ] C;China Three Gorges Univ, Coll Sci, Three Gorges Math Res Ctr, Yichang 443002, Peoples R China.
关键词:
Modified linear sampling method;variational method;interior transmission problem;partially coated dielectric
摘要:
Consider time-harmonic electromagnetic wave scattering by an infinitely long, cylindrical, orthotropic dielectric partially coated with a very thin layer of a highly conductive material, which can be modeled by a transmission problem with mixed boundary conditions. Having established the well-posedness of the direct and interior transmission problem by the variational method under certain conditions, we make use of the classical linear sampling method to reconstruct the shape of the obstacle. Then, based on a modification of the general data-to-pattern operator G, we propose a novel and simple method to justify the modified linear sampling method.
摘要:
Entities and relations extraction are the key tasks in the construction of biomedical knowledge graph, which play an important role in the biomedical artificial intelligence. However, extraction of entities and relations from biomedical texts is challenging because of the overlapping triples problem. The previous approaches typically divided the task into two separate sub-tasks. However, these methods failed to address the error propagation problem. Recent methods have been proposed to perform both sub-tasks simultaneously. Nonetheless, most current methods still encounter issues related to imbalanced interactions and independent features. In this paper, we propose a novel method based on feature partition encoding and relative positional embedding to joint extract biomedical entity and relation triples simultaneously. Compared to previous work, our method shows exceptional accurate in extracting entities and relations, while efficiently tackling the challenge of overlapping triples in biomedical texts. Our work has two contributions. Firstly, our method divides the features into task-specific and shared parts through entity, relation and sharing partitions at the encoding stage. And the encoded features will be aggregated according to the subsequent tasks. Secondly, we introduce a relative positional embedding method to capture the relative distance information between token pairs. In this way, our method can effectively deal with the sub-tasks interactions problem and improve entities and relations extraction. The experimental results show that our method improves the F1 scores of relations extraction by 3.2%, 2.1%, 3.4%, and 2.8% on four biomedical datasets, respectively.
期刊:
Science of The Total Environment,2024年922:171299 ISSN:0048-9697
通讯作者:
Chen, Mingqing
作者机构:
[Xu, Ke; Jiang, Mengling; Wang, Shuxin; Han, Qi; Gao, Xiao; Du, Wanting; Wang, Yunyi] Hubei Key Laboratory of Genetic Regulation and Integrative Biology, School of Life Sciences, Central China Normal University, Wuhan 430079, Hubei, China;[Chen, Mingqing] Hubei Key Laboratory of Genetic Regulation and Integrative Biology, School of Life Sciences, Central China Normal University, Wuhan 430079, Hubei, China. Electronic address: chenmq@mail.ccnu.edu.cn
通讯机构:
[Chen, Mingqing] H;Hubei Key Laboratory of Genetic Regulation and Integrative Biology, School of Life Sciences, Central China Normal University, Wuhan 430079, Hubei, China. Electronic address:
摘要:
After aging in the environment, some nanoplastics will carry different charges and functional groups, thereby altering their toxicological effects. To evaluate the potential impact of aging of nanoplastics on the mammalian reproductive system, we exposed C57BL/6 male mice to a dose of 5mg/kg/d polystyrene nanoparticles (PS-NPs) with different functional groups (unmodified, carboxyl functionalized and amino functionalized) for 45days for this study. The results suggest that PS-NPs with different functional groups triggered oxidative stress, a decreased in the testis index, disruption of the outer wall of the seminiferous tubules, reduction in the number of spermatogonia cells and sperm counts, and an increased in sperm malformations. We performed GO and KEGG enrichment analysis on the differentially expressed proteins, and found they were mainly enriched in protein transport, RNA splicing and mTOR signaling. We confirmed that the PI3K-AKT-mTOR pathway is over activated, which may lead to reduction of spermatogonia stem cells by over differentiation. Strikingly, PS-NPs with functional group modifications are more toxic than those of unmodified polystyrene, and that PS-NPs with positively charged amino modifications are the most toxic. This study provides a new understanding for correctly evaluating the toxicological effects of plastic aging, and of the mechanism responsible for the reproductive toxicity caused by nanoplastics.
期刊:
FRONTIERS IN PSYCHOLOGY,2024年15:1330316 ISSN:1664-1078
通讯作者:
Xu, Q
作者机构:
[Xue, Yan] Cent China Normal Univ, Fac Artificial Intelligence Educ, Wuhan, Peoples R China.;[Xu, Quan; Xu, Q] Cent China Normal Univ, Sch Foreign Languages, Wuhan, Peoples R China.
通讯机构:
[Xu, Q ] C;Cent China Normal Univ, Sch Foreign Languages, Wuhan, Peoples R China.
摘要:
Background Previous studies have examined the gender and geographic diversity within editorial boards across various academic disciplines, excluding the field of education. Thus, the purpose of this study was to address this gap by investigating the extent of gender and geographic disparity within the editorial boards of education journals.Methods The selection of top five education journals from each quartile (Q1-Q4) was performed based on Clarivate Analytics' Journal Citation Reports (JCR) 2021 within the category of "Education & Educational Research." The information of editors was collected through online sources.Results Overall, female editors accounted for 17 out of the 29 editors-in-chief (58.62%), 36 out of the 64 deputy/associate editors (56.25%), 378 out of the 728 editorial/advisory board members (51.92%) and 15 out of the 28 other types of editors (53.57%). There was no significant association between the impact factor (IF) and the proportion of female editors (Pearson's r = -0.095, p = 0.689). The United States had the highest number of editors (n = 459, 54.06%), followed by the United Kingdom (n = 98, 11.54%), Australia (n = 63, 7.42%), China (n = 29, 3.42%), Germany (n = 25, 2.94%), and Canada (n = 22, 2.59%). Notably, the majority of the included editors were from developed countries (n = 794, 93.52%), while editors from developing countries constituted a significantly smaller proportion (n = 55, 6.48%).Conclusion Editorial boards of education journals exhibit reasonable gender diversity than other disciplines, though still fall short when considering the proportion of women in the discipline. Besides, obvious geographical disparity was observed among editorial boards of education journals. There was a notable lack of representation of researchers associated with institutions from developing countries on the editorial boards of education journals. While maintaining sufficient gender diversity, it is imperative to enhance the geographical diversity in these journals, ensuring a more equitable number of positions to individuals from these underrepresented groups.
摘要:
Abstract Background Leaderboards are a highly popular gamification component used in student learning to enhance motivation, attentional engagement, and learning performance. However, few studies have examined the effects of individual leaderboard elements on English vocabulary learning through video lectures. Objectives The present study aimed to examine how different leaderboard elements (i.e., points and rank) may affect students' English vocabulary learning through video lectures. Methods A total of 34 students were assigned to groups using different leaderboard elements in a counterbalanced order. Participants' motivation, eye movements, and learning performance were measured and analysed. Results and Conclusions Students' leaderboard rank was shown to increase their motivation regardless of whether other elements were present. Eye movement tracking revealed that the presence of the leaderboard increased students' saccades between the questions and the options, and lengthened their dwell time on the learning materials while reducing their dwell time on the non‐learning‐related screen areas. Presenting students' rank alone also improved their learning performance. Implications Our findings strongly support the use of video lectures for English vocabulary learning, with the following recommendations: (1) Instructors should present students' rank on the leaderboard to enhance students' motivation and engagement; (2) Instructors should present only the students' rank on the leaderboard to also enhance students' learning performance.
期刊:
British Journal of Educational Technology,2024年 ISSN:0007-1013
通讯作者:
Wang, Zhifeng;Luo, H
作者机构:
[Liao, Xiaofang; Luo, H; Wang, Zhifeng; Luo, Heng] Cent China Normal Univ, Fac Artificial Intelligence Educ, Wuhan 430079, Peoples R China.;[Zhang, Xuedi] South China Normal Univ, Zengcheng Sch, High Sch, Guangzhou, Peoples R China.
通讯机构:
[Luo, H ; Wang, ZF] C;Cent China Normal Univ, Fac Artificial Intelligence Educ, Wuhan 430079, Peoples R China.
摘要:
Abstract Formative assessment is essential for improving teaching and learning, and AI and visualization techniques provide great potential for its design and delivery. Using NLP, cognitive diagnostic and visualization techniques designed to analyse and present students' monthly exam data, we developed an AI‐enabled visual report tool comprising six modules and conducted an empirical study of its effectiveness in a high school biology classroom. A total of 125 students in a ninth‐grade biology course were assigned to a treatment group (n = 63) receiving AI‐enabled visual reports as the intervention and a control group (n = 62) receiving overall oral feedback from the teacher. We present the main statistical results of the within‐subjects design and the between‐subjects design respectively, to better capture the main findings. Repeated measures ANOVA revealed a significant interaction effect of intervention and time on learning achievement, and the paired‐sample Wilcoxon test indicated that the treatment group had experienced increasing learning anxiety (Cohen's d = 0.203, p = 0.046) and self‐efficacy (Cohen's d = 1.793, p = 0.000) over time. Moreover, we conducted a series of non‐parametric tests to compare the effects of AI‐enabled visual reports and teacher feedback, but found no significant differences except for an increased self‐efficacy (Cohen's d = 0.312, p = 0.046). Additionally, we had the students in the treatment group rate their favourable modules in the AI‐enabled visual report and provide evaluative feedback. The study results provide important insights into the design and implementation of effective formative assessment supported by artificial AI and visualization techniques. Practitioner notes What is already known about this topic Formative assessment is essential for improving teaching and learning. Traditional formative assessment tools lack accurate data‐oriented assessment and usability. AI and visualization techniques have great potential for formative assessment. What this paper adds This study designs and implements an AI‐enabled visual report tool that generates data‐driven, user‐friendly reports. The AI‐enabled visual report can not only enhance students' learning achievement and self‐regulated learning over time but also increase their test anxiety. The AI‐enabled visual report has a comparable effect with teacher feedback but leads to increased self‐efficacy. Implications for practice and/or policy We recommend using the AI‐enabled visual report in large‐size classes for its overall positive effects on both learning achievement and self‐regulated learning. We recommend using the AI‐enabled visual report over teacher feedback for its capacity to enhance students' self‐efficacy. We recommend prioritizing the modules of Performance Ranking, Personal Mastery and Knowledge Alert when designing the AI‐enabled visual report.
作者机构:
[Guan, Rui; Sun, Yao; Li, Junrong; Liu, Guorong] Cent China Normal Univ, Coll Chem, Int Joint Res Ctr Intelligent Biosensor Technol &, Natl Key Lab Green Pesticide, Wuhan 430079, Peoples R China.;[Yan, Mingzhe] Huazhong Univ Sci & Technol, Wuhan Jinyintan Hosp, Tongji Med Coll, Wuhan 430023, Peoples R China.;[Cheng, Jing] Wuhan Univ Sci & Technol, Sch Publ Hlth, Wuhan 430072, Peoples R China.;[Zhan, Jianbo] Hubei Prov Ctr Dis Control & Prevent, Inst Hlth Inspect & Testing, Wuhan 430072, Peoples R China.;[Wuethrich, Alain; Trau, Matt; Wuethrich, A] Univ Queensland, Australian Inst Bioengn & Nanotechnol, Ctr Personalized Nanomed, Brisbane, Qld 4072, Australia.
通讯机构:
[Wuethrich, A ] U;[Sun, Y ] C;Cent China Normal Univ, Coll Chem, Int Joint Res Ctr Intelligent Biosensor Technol &, Natl Key Lab Green Pesticide, Wuhan 430079, Peoples R China.;Univ Queensland, Australian Inst Bioengn & Nanotechnol, Ctr Personalized Nanomed, Brisbane, Qld 4072, Australia.
摘要:
The molecular detection of multiple respiratory viruses provides evidence for the rational use of drugs and effective health management. Herein, we developed and tested the clinical performance of an electrohydrodynamic-driven nanobox-on-mirror platform (E-NoM) for the parallel, accurate, and sensitive detection of four respiratory viral antigens. The E-NoM platform uses gold-silver alloy nanoboxes as the core material with the deposition of a silver layer as a shell on the core surfaces to amplify and enable a reproducible Raman signal readout that facilitates accurate detection. Additionally, the E-NoM platform employs gold microelectrode arrays as the mirror with electrohydrodynamics to manipulate the fluid flow and enhance molecular interactions for an improved biosensing response. The presence of viral antigens binds the nanobox-based core-shell nanostructure on the gold microelectrode and creates the nanocavity with extremely strong "hot spots" to benefit sensitive analysis. Significantly, in a large clinical cohort with 227 patients, the designed E-NoM platform demonstrates the capability of screening respiratory infection with achieved clinical specificity, sensitivity, and accuracy of 100.0, 96.48, and 96.91%, respectively. It is anticipated that the E-NoM platform can find a position in clinical usage for respiratory disease diagnosis.
作者机构:
[Li, Qian; Yan, Qiang; Yan, Q; Li, Yuntao] Zhejiang Univ, Huzhou Cent Hosp, Affiliated Huzhou Hosp, Dept Gen Surg,Sch Med, Huzhou 313000, Peoples R China.;[Li, Qian; Sun, Yao; Zhao, Fang] Cent China Normal Univ, Coll Chem, Natl Key Lab Green Pesticide, Wuhan 430079, Peoples R China.;[Ye, Huan] Guangxi Univ, Sch Light Ind & Food Engn, Nanning 530004, Peoples R China.;[Zhang, ZP; Zhang, Zhipeng] Hubei Univ Sci & Technol, Xianning Med Coll, Coll Pharm, Xianning 437100, Peoples R China.
通讯机构:
[Zhang, ZP ] H;[Sun, Y ] C;[Yan, Q ] Z;Zhejiang Univ, Huzhou Cent Hosp, Affiliated Huzhou Hosp, Dept Gen Surg,Sch Med, Huzhou 313000, Peoples R China.;Cent China Normal Univ, Coll Chem, Natl Key Lab Green Pesticide, Wuhan 430079, Peoples R China.
摘要:
Bacterial infections can lead to the development of large-scale outbreaks of diseases that pose a serious threat to human life and health. Also, conventional antibiotics are prone to producing resistance and allergic reactions, and their therapeutic effect is dramatically diminished when bacterial communities form biofilms. Fortunately, well-designed supramolecular coordination complexes (SCCs) have been used as antibacterials or anti-biofilms in recent years. SCCs can kill bacteria by directly engaging with the bacterial surface through electrostatic interactions or by penetrating the bacterial membrane through the auxiliary effect of cell-penetrating peptides. Furthermore, scientists have engineered fluorescent SCCs that can produce reactive oxygen species (ROS) to eliminate bacteria when exposed to laser irradiation, and they also demonstrate outstanding performance in in vivo imaging, enabling integrated diagnosis and treatment. In this review, we summarize the design strategy and applications of SCCs in antibacterials or anti-biofilms and provide an outlook on future research.
期刊:
Sustainable Energy, Grids and Networks,2024年38:101293 ISSN:2352-4677
通讯作者:
Yi Xiao
作者机构:
[Yi Xiao; Sheng Wu; Chen He; Ming Yi] School of Information Management, Central China Normal University, Wuhan 430079, PR China;[Yi Hu] School of Economics and Management, University of Chinese Academy of Sciences, Beijing 100190, PR China
通讯机构:
[Yi Xiao] S;School of Information Management, Central China Normal University, Wuhan 430079, PR China
摘要:
The random and fluctuating nature of wind energy brings tremendous challenges and disturbances to the security operation of wind power systems, accurate wind power prediction can effectively reduce these negative impacts. To this end, this paper proposes a hybrid wind power prediction model based on the "decomposition-reconstruction-ensemble" strategy, which consists of four main components, namely decomposition, reconstruction, prediction, and ensemble. Specifically, the original wind power series is decomposed into several sub-modes and reconstructed by frequency by the sample entropy(SE)-optimized variational modal decomposition(VMD) algorithm, subsequently, the Pearson correlation coefficients between the wind speed time series and the reconstructed components of wind power are calculated to divide the wind power series into trend and fluctuation components. Then both the two components are sequentially predicted using the temporal convolutional network(TCN) model. The final predicted value is obtained from the set of predicted results for each component. The wind power data from two wind farms in Hami, Xinjiang are adopted as examples for empirical study, and the results show that the IVMD-R-TCN model proposed in this paper performs significantly better than the benchmark model, which illustrates the predictive validity of the proposed model and is an effective tool for wind power forecasting.
作者机构:
[Yang, Xiao-Chen] College of Biology and Environmental Sciences, Jishou University, Jishou 416000 , Hunan Province, China;[Lv, Xiao-Wen; Huang, Shuang-Quan] Institute of Evolution and Ecology, School of Life Sciences, Central China Normal University , Wuhan 430079, China;Department of Ecology, College of Life Sciences, Wuhan University , Wuhan 430072, China;[Feng, Hui-Hui] Institute of Evolution and Ecology, School of Life Sciences, Central China Normal University , Wuhan 430079, China<&wdkj&>Department of Ecology, College of Life Sciences, Wuhan University , Wuhan 430072, China
通讯机构:
[Shuang-Quan Huang] I;Institute of Evolution and Ecology, School of Life Sciences, Central China Normal University , Wuhan 430079, China
摘要:
BACKGROUNDS AND AIMS: The hypothesis that plants evolve features that protect accessible pollen from consumption by flower visitors remains poorly understood. METHODS: To explore potential chemical defenses against pollen consumption, we examined the pollinator assemblage, foraging behaviour, visitation frequency and pollen transfer efficiency in Rhododendron molle, a highly toxic shrub containing Rhodojaponin III. Nutrient (protein and lipid) and toxic components in pollen and other tissues were measured. KEY RESULTS: Overall in the five populations, floral visits by butterflies and bumblebees were relatively more frequent than visits by honeybees. All foraged for nectar but not pollen. Butterflies did not differ from bumblebees in the amount of pollen removed per visit, but deposited more pollen per visit. Pollination experiments indicated that R. molle was self-compatible, but both fruit and seed production were pollen limited. Our analysis indicated that the pollen was not protein-poor and had a higher concentration of the toxic compound Rhodojaponin III than petals and leaves, which compound was undetectable in nectar. CONCLUSION: Pollen toxicity in Rhododendron flowers may discourage pollen robbers (bees) from taking the freely accessible pollen grains, while the toxin-free nectar rewards effective pollinators, promoting pollen transfer. This preliminary study supports the hypothesis that chemical defense in pollen would be likely to evolve in species without physical protection from pollinivores.
作者机构:
[Chen, Jingying; Wang, Zhun; Xu, Ruyi; Zhou, Longpu] Cent China Normal Univ, Fac Artificial Intelligence Educ, Wuhan 430079, Hubei, Peoples R China.;[Zhou, Longpu] Ningbo Yuxing Educ Technol Co Ltd, Ningbo, Zhejiang, Peoples R China.
通讯机构:
[Chen, JY ] C;Cent China Normal Univ, Fac Artificial Intelligence Educ, Wuhan 430079, Hubei, Peoples R China.
关键词:
Facial expression intensity estimation;Ordinal regression;Label distribution learning;Semi-supervised
摘要:
Facial expression intensity estimation has promising applications in health care and affective computing, such as monitoring patients’ pain feelings. However, labeling facial expression intensity is a specialized and time-consuming task. Ordinal regression (OR)-based methods address this issue to some extent by estimating the relative intensity but failing to estimate the absolute intensity due to lack of exploring useful information from noisy labels caused by manual and automatic labeling biases. Inspired by label distribution learning (LDL) to resist the noisy labels, this paper introduces the label-distribution-learning-enhanced OR (LDL-EOR) approach for facial expression intensity estimation. This design aims to utilize LDL to improve the accuracy of absolute intensity estimation while keeping the cost of manual labeling low. The label distribution is converted into a continuous intensity value by calculating the mathematical expectation, which makes the prediction results meet both relative and absolute intensity constraints. Ensuring the feasibility of LDL-EOR in different supervised settings, this paper presents a unified label distribution generation framework to automatically relabel training data frame by frame. The generated soft labels are used to supervise the LDL-EOR model and enhance its robustness to the noise existing in the original labels. Numerous experiments were conducted on three public expression datasets (CK+, BU-4DFE, and PAIN) to validate the superiority of LDL-EOR relative to other state-of-the-art approaches.