作者机构:
[Xu B.; Zong J.F.] School of Computer Science and Technology, Hubei University of Science and Technology, Xianning, 437100, China;[Xu Z.] School of Educational Information Technology, Central China Normal University, Wuhan, 430072, China;[Shu C.; Xiao J.] School of Electronic Information, Wuhan University, Wuhan, 430072, China;[Ding L.] School of Computer Science and Technology, Hubei University of Science and Technology, Xianning, 437100, China, School of Electronic Information, Wuhan University, Wuhan, 430072, China
会议名称:
1st International Symposium on Geometry and Vision, ISGV 2021
会议时间:
28 January 2021 through 29 January 2021
关键词:
Convolutional neural network;Detection;Dotted line;The lane line;The solid line
作者机构:
Authors to whom correspondence should be addressed.;School of Information Technology in Education, South China Normal University, Guangzhou 510631, China;[Qingtang Liu; Linjing Wu] School of Educational Information Technology, Central China Normal University, Wuhan 430079, China;[Yunxiang Zheng; Jingxiu Huang] Authors to whom correspondence should be addressed.<&wdkj&>School of Information Technology in Education, South China Normal University, Guangzhou 510631, China
通讯机构:
[Yunxiang Zheng; Jingxiu Huang] A;Authors to whom correspondence should be addressed.<&wdkj&>School of Information Technology in Education, South China Normal University, Guangzhou 510631, China
关键词:
comma disambiguation;feature engineering;hyperparameter tuning;imbalanced learning;natural language understanding;random forests
摘要:
Natural language understanding technologies play an essential role in automatically solving math word problems. In the process of machine understanding Chinese math word problems, comma disambiguation, which is associated with a class imbalance binary learning problem, is addressed as a valuable instrument to transform the problem statement of math word problems into structured representation. Aiming to resolve this problem, we employed the synthetic minority oversampling technique (SMOTE) and random forests to comma classification after their hyperparameters were jointly optimized. We propose a strict measure to evaluate the performance of deployed comma classification models on comma disambiguation in math word problems. To verify the effectiveness of random forest classifiers with SMOTE on comma disambiguation, we conducted two-stage experiments on two datasets with a collection of evaluation measures. Experimental results showed that random forest classifiers were significantly superior to baseline methods in Chinese comma disambiguation. The SMOTE algorithm with optimized hyperparameter settings based on the categorical distribution of different datasets is preferable, instead of with its default values. For practitioners, we suggest that hyperparameters of a classification models be optimized again after parameter settings of SMOTE have been changed.
作者机构:
[Hu Z.] School of Educational Information Technology, Central China Normal University, Wuhan, China;[Su J.] School of Computer Science, Hubei University of Technology, Wuhan, China;[Koroliuk Y.] Chernivtsi Institute of Trade and Economics, Kyiv National University of Trade and Economics, Chernivtsi, Ukraine
会议名称:
3rd International Conference on Computer Science, Engineering and Education Applications, ICCSEEA 2020
会议时间:
21 January 2020 through 22 January 2020
关键词:
Collaborative learning;Interactive learning environments;Predicting of academic performance;Simulations;Teaching strategies
作者机构:
[Gang, Zhao; Qing, Xia; Biling, Hu; Jie, Chu; Wenjuan, Zhu; Hui, He] Cent China Normal Univ, Fac Artificial Intelligence Educ, Sch Educ Informat Technol, Wuhan, Peoples R China.
通讯机构:
[Zhu Wenjuan] S;School of Educational Information Technology, Faculty of Artificial Intelligence Education, Central China Normal University, Wuhan, China
关键词:
Teacher behavior;Behavior recognition;Motion region extraction;Key-points tracking;Teacher’s behavior rule
摘要:
The analysis of teacher behavior of massive teaching videos has become a surge of research interest recently. Traditional methods rely on accurate manual analysis, which is extremely complex and time-consuming for analyzing massive teaching videos. However, existing works on action recognition are difficultly transplanted to the teacher behavior recognition, because it is difficult to extract teacher’s behavior from complex teaching scenario, and teacher’s behaviors are given professional educational semantics. These methods are not adequate for the need of the teacher behavior recognition. Thus, a novel and simple recognition method of teacher behavior in the actual teaching scene for massive teaching videos is proposed, which can provide technical assistance for analyzing teacher behavior and fill the blank of automatic recognition of teacher behavior in actual teaching scene. Firstly, we discover the educational pattern which it be named “teacher set”, that is, the spatial region of the video of the whole class where teachers should exist. Based on this, the algorithm of teacher set identification and extraction (Teacher-set IE algorithm) is studied to identify the teacher in the teaching video, and reduce the interference factors of classroom background. Then, an improved behavior recognition network based on 3D bilinear pooling (3D BP-TBR) is presented to enhance fusion representation of three-dimensional features thus identifying the categories of teacher behavior, and experiments show that 3D BP-TBR can achieve better performance on public and self-built dataset (TAD-08). Hence, our whole approach can increase recognition accuracy of teacher behavior in the actual teaching scene to utilize the deep integration of educational characteristics and action recognition technology.
摘要:
In the electronic age, the changes in the mirror space constructed by movies are also brought about by changes in the media. Cinematic space is a four-dimensional space illusion including time created by the use of light and shadow, color, perspective, sound, and the movement of characters and cameras. It is not a real four-dimensional space, but a reflection and reproduction of the real space. With the goal of improving the accuracy of spatially coded structured light 3D reconstruction, this paper conducts in-depth research on several key technologies that affect the reconstruction accuracy and makes corresponding innovations and improvements. Moreover, this paper respectively proposes an adaptive structured light spatial coding algorithm based on geometric features and a color structured light decoding algorithm based on color shift technology. In addition, this paper implements a spatially coded structured light three-dimensional reconstruction system and calibration system. The simulation research shows that the method in this paper has certain reliability.
作者机构:
[Nie, Yanjiao; Luo, Heng] Cent China Normal Univ, Sch Educ Informat Technol, Wuhan, Peoples R China.;[Sun, Di] Syracuse Univ, Dept Instruct Design Dev & Evaluat, Syracuse, NY USA.
通讯机构:
[Luo, Heng] C;Cent China Normal Univ, Sch Educ Informat Technol, Wuhan, Peoples R China.
关键词:
Massive open online course;diagnostic evaluation;Analytic Hierarchy Process;emotion classification;quality assurance
摘要:
Object co-segmentation is a challenging task, which aims to segment common objects in multiple images at the same time. Generally, common information of the same object needs to be found to solve this problem. For various scenarios, common objects in different images only have the same semantic information. In this paper, we propose a deep object co-segmentation method based on channel and spatial attention, which combines the attention mechanism with a deep neural network to enhance the common semantic information. Siamese encoder and decoder structure are used for this task. Firstly, the encoder network is employed to extract low-level and high-level features of image pairs. Secondly, we introduce an improved attention mechanism in the channel and spatial domain to enhance the multi-level semantic features of common objects. Then, the decoder module accepts the enhanced feature maps and generates the masks of both images. Finally, we evaluate our approach on the commonly used datasets for the co-segmentation task. And the experimental results show that our approach achieves competitive performance. (C) 2020 Elsevier B.V. All rights reserved.
作者:
Hongxia Li;ChengLing Zhao*;Taotao Long;Yan Huang;Fengfang Shu
期刊:
British Journal of Educational Technology,2021年52(6):2263-2277 ISSN:0007-1013
通讯作者:
ChengLing Zhao
作者机构:
[Hongxia Li; ChengLing Zhao; Taotao Long; Yan Huang; Fengfang Shu] School of Educational Information Technology, Central China Normal University, Wuhan, China
通讯机构:
[ChengLing Zhao] S;School of Educational Information Technology, Central China Normal University, Wuhan, China
期刊:
International Journal of Human-Computer Interaction,2021年37(9):884-901 ISSN:1044-7318
通讯作者:
Liu, Shan;Zhu, Wen-Juan
作者机构:
[Qi, Yu-Heng; Liu, Shan; Zhu, Wen-Juan; Liu, S; Zhao, Gang] Cent China Normal Univ, Sch Educ Informat Technol, Wuhan 430079, Peoples R China.
通讯机构:
[Liu, S; Zhu, WJ] C;Cent China Normal Univ, Sch Educ Informat Technol, Wuhan 430079, Peoples R China.
摘要:
Mobile augmented reality (AR) technology creates realistic learning situations and a strong sense of immersion, which is conducive to enhance learning experience and stimulate learning motivation. However, existing mobile outdoor augmented reality applications generally have a complicated operation process and a mismatch between learning resources and corresponding scenes, which leads to a poor learning experience. Therefore, we propose a lightweight mobile outdoor AR method that combines deep learning and knowledge modeling to perceive learning scenes with a goal to improve learning experience. This method improves the accuracy of scene perception and resources retrieval and provides a convenient mobile AR technology solution for outdoor learning. To evaluate the proposed method, we provide objective criteria to assess the effectiveness of the lightweight object detection model and the learning resources retrieval approach. Simultaneously, we investigate the evaluation of participants majoring in teacher education on the usability of the proposed method by the modified system usability scale questionnaire and net promoter score. Experimental results demonstrate that our method achieves high detection accuracy, good usability, and is of great significance to improve outdoor learning experience.
作者机构:
[Xu, P.; Zhang, M.] Cent China Normal Univ, Fac Artificial Intelligence Educ, Sch Educ Informat Technol, Wuhan 430079, Peoples R China.;[Xu, P.; Zhang, M.] Cent China Normal Univ, Natl Language Resources Monitoring & Res Ctr Netw, Wuhan 430079, Peoples R China.;[Xu, P.] Hubei Meteorol Serv Ctr, Wuhan 430079, Peoples R China.
通讯机构:
[Zhang, M.] C;Cent China Normal Univ, Fac Artificial Intelligence Educ, Sch Educ Informat Technol, Wuhan 430079, Peoples R China.;Cent China Normal Univ, Natl Language Resources Monitoring & Res Ctr Netw, Wuhan 430079, Peoples R China.
关键词:
node attribute;generative adversarial network;network embedding;policy gradient;bidirectional long short term memory
摘要:
Graph generative adversarial network has achieved remarkable effectiveness, such as link prediction, node classification, user recommendation and node visualization in recent years. Most existing methods mainly focus on how to represent the proximity between nodes according to the structure of the graph. However, the graph nodes also have rich attribute information in social networks, the traditional methods mainly consider the node attributes as auxiliary information incorporate into the embedding representation of the graph to improve the accuracy of node classification and link prediction. In fact, in social networks, these node attributes are often sparse. Due to privacy and other reasons, the attributes of many nodes are difficult to obtain. Inspired by the application of generative adversarial network in image field, we propose an innovative framework to discover node latent attribute. Through experiments, we demonstrate the effectiveness of our proposed methods.