Mobile QR Code QR CODE

2024

Acceptance Ratio

21%


  1. (School of Art, Wuhan Business University, Wuhan, 430056, China xiaopeng_wbu@163.com)



iOS, art, Information entropy, Online teaching, Attention mechanism

1. Introduction

With the general improvement of modern people's living standards, cultivating one's cultural literacy and hobbies in daily life has gradually become a modern life trend. In addition to participating in external cultural life, cultivating one's own cultural creativity is also a way of cultural life, especially in the field of art. However, due to the fact that most people only receive relatively basic art education, they lack the foundation of advanced art aesthetics and painting. In the process of self-learning, the lack of guidance and understanding of oneself often leads to difficulties in painting, making it difficult to continue art learning [1,2,3]. Therefore, it is necessary to have an art teaching system that can provide portable and interactive guidance to the beginners at any time. The difference between beginner and mature painting instruction lies in the former's ability to guide newcomers to establish their own artistic style and direction. Achieving this requires initial classification and positioning of a beginner's painting style. By following this approach, students can determine their future artistic career path based on their current foundation and generate more well-planned inquiries [4,5,6]. However, planning alone is insufficient. Effective communication among students is also imperative in an interactive art learning environment. The IOS platform offers a seamless mode for real-time human-computer and human-to-human interaction, providing an optimal solution. Applying computer algorithm technology to modern art teaching can effectively enhance the intelligence of modern art teaching, making it more universal and user-friendly, and also more efficient and personalized. On the other hand, it can also enhance the utilization of computer algorithms and establish a basis for further enhancement, optimization, and expansion of their applications in various fields [7,8,9].

2. Related Works

The team of Groenendijk designed an art education assessment system for teachers and students in secondary art education. The system is divided into two parts: teacher evaluation and student evaluation, which helps teachers and students communicate better in the process of art teaching and learning [10]. Simamora's team conducted research on the effects of the COVID-19 pandemic on higher art education and piloted an art teaching method using both synchronous and asynchronous online education. The study utilized both qualitative and quantitative analysis methods. The study's findings indicate that art education in Xi'an is an effective teaching method that satisfies instructional objectives [11]]. Long designed an art teaching achievement evaluation model that uses the ID3 decision tree algorithm and incorporates data flow technology to comprehensively evaluate students' learning outcomes based on their grades in art exams. The experimental findings suggest that the algorithm created in this research has high practical value and a beneficial impact on the teaching of art [12]. Gao designed an efficient hybrid education strategy for art and design courses. From an engineering application perspective, the study develops a hybrid teaching performance assessment method. This method is specifically designed for art and design courses. Additionally, an evaluation analysis model is established as part of this study. These tools can quantify the art teaching performance. Thus, they provided a more effective practical solution for teaching [13]. Miller has integrated STEAM methods into art curriculum teaching to create effective and appropriate professional development opportunities for art teachers and students. He integrated artistic concepts into the types of classroom teaching and provided creativity for the classroom [14].

The paper utilized information entropy technology and attention mechanism during model design, which has seen significant development in recent years. Zhang L's team employed information entropy technology and elasticity method to evaluate the quality index system of the regional ecological environment, examining the impact of land use trajectory on ecological surroundings. Environmental changes have caused inconsistencies between the degree of land use in certain areas and the resulting changes in ecological environment quality. Additionally, a recent study indicates that irrational land use is more likely to contribute to the degradation of the ecological environment [15]. Wu's team applied the information entropy technique to the data generation and visualization of highlight map images, and designed a strategy to judge the level of highlight map quality from the perspective of image information and image definition. The designed model can be used to judge the quality of the highlight spectrum. It also can objectively form visualization results from batches of hyperspectral image species, and the method can be used to evaluate different types of dimensionality reduction algorithms [16]. Fukui, et al. integrated attentional mechanisms with convolutional neural networks to enhance the scalability of visual interpretation models. They demonstrated that introducing attentional mechanisms enabled end-to-end training for multiple image recognition tasks. The results show that this model is superior to the baseline model in image recognition [17]. Zhu conducted a study on the attention mechanism that took a non-traditional approach, which involved the removal of various spatial attention elements. This approach resulted in significant findings regarding spatial attention in deep networks. The study demonstrates that utilizing a proper fusion of variable convolutional neural networks and key content saliency can attain optimal accuracy and efficiency balance [18]. Xiong presented a neural network architecture in the form of a graph, developed around a molecular-based representation of attention mechanisms for visualizing focused features from a graphical attention perspective. The model is capable of learning molecular interactions for a specific task, as demonstrated by the results [19].

This paper summarizes the research results of multiple teams in the field of art education. Among them, some teams focus on designing evaluation systems and blended education strategies to improve communication and learning outcomes between teachers and students in the art teaching process. Other teams are integrating STEAM methods into art courses and incorporating art concepts into classroom teaching to offer more effective creative solutions. Additionally, some teams use information entropy technology and attention mechanisms to evaluate environmental quality assessment index systems and hyperspectral image quality. While these studies present beneficial tools and technologies for art education, they contain shortcomings in research techniques, data sources, and sample selection. Hence, forthcoming research should focus on enhancing the design and analysis of empirical research, and utilizing more representative samples to substantiate results. At the same time, it is also necessary to consider how to integrate and compare different research results to obtain a more comprehensive and in-depth understanding.

This study combines information entropy techniques with attention mechanisms and applies them to art teaching to provide a new approach to modern interactive art education.

3. Ios interactive art teaching system design

3.1 Design of ios painting image style classification model for painting teaching

Beginners often lack a solid foundation in drawing and painting techniques, resulting in their works lacking a distinct style. Challenges arise in objectively assessing students' progress in online modern art teaching. Consequently, they encounter difficulties in receiving specific guidance and training, which hampers effective teaching interaction. Consequently, the development of students' art skills may be impeded [20,21,22]. To address this issue, a painting style classification model and a painting description generation model are designed using the portable platform IOS. It is used to evaluate students' works in the process of painting teaching, providing targeted support for teaching interaction during the learning process. The style classification model is used to locate the direction of students' future painting practice. The drawing description generation model can classify students with similar descriptions, making it convenient for students to communicate with each other during the learning process.

The design of the painting image style classification model mainly uses image information entropy technology to classify image art styles. The color entropy and block entropy contained in the participants' works are analyzed for the distribution of painting colors and lines. Based on the image information entropy characteristics, the study classifies the participants' painting images into four basic categories, which are simple line drawing category, complex line drawing category, conservative style category and exaggerated style category. Simple line painting refers to the painting of students who lack a coloring foundation and have simple line composition. Complex line painting refers to student paintings that lack a color foundation but have complex and dense line compositions. The conservative style category refers to students' paintings that use monotonous colors, even with colder tones transitioning between color blocks. Exaggerated style refers to the use of colors in students' paintings that are more exaggerated, without too many color blocks, and the overall color system tends to be warm, as shown in Fig. 1.

Fig. 1. Classification of painting styles.

../../Resources/ieie/IEIESPC.2025.14.1.11/image1.png

It is worth noting that it is often difficult for students to notice the objective characteristics contained in their own works during the drawing process. So it is necessary to classify the students' drawings and to conduct targeted teaching interaction. The study uses the IOS programming language to build a resource library of students' drawings. The specific construction is shown in Fig. 2.

Fig. 2. Resource library construction.

../../Resources/ieie/IEIESPC.2025.14.1.11/image2.png

The research focus of the iOS programming section is on the Objective-C programming language. It originated from the C language and is divided into four key features: message mechanism, protocol, category, and memory management method. The message mechanism is a mechanism that determines the method to be called at runtime, while the execution method is determined through function calls at compile time. The protocol is a method declaration that cannot provide the implementation method itself, but can be provided through specific references to the protocol. And using this proxy protocol model is also more convenient for maintaining model code. The category is a way to add methods to the current category directly without the need of inheritance. This coding method allows developers to add custom methods to existing categories. Memory management methods are divided into two main categories, i.e. manual memory reference management and automatic memory reference management. Both mechanisms are implemented through the automatic reference counting mechanism [23,24,25]. The specific programming language features are shown in Fig. 3.

Fig. 3. Programming language features.

../../Resources/ieie/IEIESPC.2025.14.1.11/image3.png

The study mainly uses the LAB color model when performing the color entropy calculation of the college works. L denotes the luminance information of the painting. A represents the information interval between the transition of a painting from a green area to a magenta area in color. B represents the information interval between the transition of the painting from the blue area to the yellow area. Color entropy technology more accurately describes the color of an image and can fully capture the color features of the image. Therefore, it has a good classification effect in painting style classification. Since the degree of color saturation is an important criterion in the evaluation of paintings, the study is mainly conducted in HSV color mode. The S channel can be used exclusively for analyzing color saturation. Firstly, RGB images are converted into HSV images, and in this image mode, the range of all channels can be simplified within [0,1], as shown in Equation (1).

(1)
$ H_{h} \sum _{i=0}^{1}P\left(i\right) \log _{2} P\left(i\right) . $

In (1), $i$ denotes the hue index and $P$ denotes the overall probability of that hue appearing in the picture of the college painting. A weighted gray scale function is introduced as shown in (2).

(2)
$ f\left(x\right)=1+\frac{m}{z} . $

In turn, (3) can be obtained.

(3)
$ H=f\left(x\right)\cdot H_{h}. $

In (3), $H$ represents the hue channel. If the color entropy calculation only considers the hue element, the participants' painting styles are classified as highly similar. This is mainly due to the lack of saturation determination, where the same hue is interpreted as the same color. To address this limitation, the study analyzed the S channel, which is exclusively dedicated to color saturation. Equation (4) can be used to calculate the average color saturation of the college painting images.

(4)
$ S_{a} =\sum _{x=0,y=0}S_{x,y} / x_{\max } \cdot y_{\max } ,~ (0\le S\le 1 ). $

In (4), $x$ and $y$ denote the pixel indexes. $S_{x,y} $ denotes the saturation value at the first $\left(x,y\right)$ pixel, which can be used as a weighting function. The color entropy calculation formulae in HSV mode can be combined in the form of (5).

(5)
$ H_{hue} =f\left(x\right)\cdot S_{a} \cdot H_{h} . $

Since the participant's drawing image comprises not only the color component but also the composition element, evaluating and categorizing the participant's drawing based solely on color entropy would not be accurate. Therefore, additional analysis of drawing lines and composition information is required. In this part of the study, the block entropy technique is used for model construction. The two-dimensional entropy model mainly distinguishes images that obtain basic information entropy through a blocking strategy, and then constructs two-dimensional image information entropy. It achieves the effect of two-dimensional student drawing images by minimizing the interference of color information in the images. Supposing the potential segmentation boundary point of the image is $S$, and the probability condition of the image gray level is represented by $pL$, which can be divided into two probability distributions. The first probability distribution is shown in (6).

(6)
$ A:~\frac{p_{0} }{p_{d} } ,~\frac{p_{1} }{p_{d} } ,~\frac{p_{2} }{p_{d} } ,~\frac{p_{3} }{p_{d} } ,~\cdots ,~\frac{p_{s} }{p_{d} }. $

The second probability distribution is shown in (7).

(7)
$ B:~\frac{p_{s+1} }{1-p_{d} } ,~\frac{p_{s+2} }{1-p_{d} } ,~\frac{p_{l} }{1-p_{d} }. $

The $l$ in (7) indicates the number of gray levels. The entropy value after image segmentation is shown in (8).

(8)
$ H\left(S\right)=E_{A} +E_{B} . $

$E_{A} $ in (8) can be calculated in the form of (9).

(9)
$ E_{A} =\sum _{i=0}^{s}\frac{p_{i} }{p_{d} } \log \left(\frac{p_{i} }{p_{d} } \right) . $

In (10), $E_{B} $ can be calculated as the form of (10).

(10)
$ E_{B} =\sum _{i=S+1}^{L}\frac{p_{i} }{1-p_{d} } \log \left(\frac{p_{i} }{1-p_{d} } \right) . $

Then the optimal threshold of the segmented image can be expressed in the form of (11).

(11)
$ H_{0} =\arg \max \left(H\left(S\right)\right)_{S} . $

In addition, in the block entropy section, research mainly uses variance form to characterize the spatial distribution of images and distinguishes students' painting images based on specific rules. The image is divided into small blocks and the information entropy of each individual block is calculated. Finally, the information entropy variance and block grayscale distribution of all blocks are calculated to analyze the connections between the blocks. The overall architecture of the image entropy style classification model is shown in Fig. 4.

Fig. 4. Overall architecture of image entropy style classification model.

../../Resources/ieie/IEIESPC.2025.14.1.11/image4.png

3.2 Painting description generation model design

The painting description generation model utilized in this study primarily utilizes a multi-attention mechanism to generate image descriptions via a triple-attention mechanism. The specific structure is shown in Fig. 5.

Fig. 5. Structure of triple attention mechanism.

../../Resources/ieie/IEIESPC.2025.14.1.11/image5.png

As shown in Fig. 5, ATT1 represents the attention mechanism for the spatial information of the image. V represents the attention context vector of the previous moment. ATT2 represents the attention mechanism for the hidden unit of the previous moment. ATT3 represents the attention mechanism of the image information with the hidden unit of the LSTM model at the current moment. The study uses the LSTM model to generate description statements for the participant's painting. In the input stage, the weight matrix can be described as (12).

(12)
$ Z_{it} =w_{ih}^{T} \tanh \left(W_{iv} V+\left(W_{ih} h_{t-1} \right)a_{i}^{T} \right). $

In (12), $Z_{it} $ denotes the weight matrix. $W_{iv} $, $w_{ih} $, and $W_{th} $ denote the parameters obtained by the training method, respectively. $V$ denotes the feature elements of the image space. $h_{t-1} $ denotes the hidden state of the LSTM model at the previous moment. Then the mapping of the attention model in the image mapping space dimension can be expressed in the form of (13).

(13)
$ a_{it} =\max (Z_{ti} ). $

The attention mechanism can be expressed in the form of (14).

(14)
$ c_{it} =\sum _{m=1}^{k}\alpha _{itm} v_{m} . $

The design of the model first extracts data from words that can describe college painting, and then summarizes it into a database of painting description words. Next, by indexing, the word vector of each word can be set to correspond to its position in the word matrix. The attention mechanism design part adopts the method of feature vector matching to match the text feature vectors of the subject's painting with the visual features of the painting image, forming a vector matrix with correlation. The elements in this matrix are strongly correlated with each other. The study utilizes the text feature and visual feature vectors as input for the attention module, respectively. The final attention vector is then formed through a weighted sum calculation, as demonstrated in (15).

(15)
$ a_{j} =\sum _{i=1}^{N}w_{i,j} v_{i}. $

In (15), $w$ represents the output formed after the parameter matrix is input to the SoftMax layer. $v$ represents the weights of the visual feature vector.

4. Effectiveness test of art teaching system

4.1 Analysis of the effect of painting image style classification

When testing the effectiveness of painting image style classification in this study, it is first necessary to determine the sample size, type labels for testing applications, and testing indicators for experimental applications, as shown in Table 1.

Table 1. Experimental setup.

Test elements

Type or value of test application

Image of participant's drawing

800 sheets

Single type of participant drawing images

200 sheets

Types of images drawn by participants

4 categories

Test type label name

Conservative and simple

Exaggerated simplicity

Conservative and complex

Exaggerated complexity

Test application index

Accuracy

Recall Rate

Table 1 outlines the experiment designed for this study, which comprises 800 participant drawings categorized into four types, each containing 200 images. The labels used in the experiment are classified according to four basic categories commonly found among participants: simple drawing, complex drawing, conservative style, and exaggerated style. Ultimately, four types of labels are formed: conservative simplicity, exaggerated simplicity, conservative complexity, and exaggerated complexity. The study utilizes two primary indexes, accuracy and recall, to evaluate the efficacy of the application. For the model training phase, 100 images from each category in the painting image collection are selected as the training set, with the remaining 100 used as the test set. The accuracy results of the model are shown in Fig. 6.

Fig. 6. Accuracy results.

../../Resources/ieie/IEIESPC.2025.14.1.11/image6.png

In Fig. 6, the number of style classification results of the designed painting image style classification model is 58 for conservative simple types, 31 for exaggerated types, 46 for conservative types, and 55 for exaggerated complex types. The conservative simple type has the highest number of classifications, while the exaggerated simple type has the lowest number of classifications. In terms of the correct classification number of military academy students' paintings, the conservative simple type is 51, the exaggerated simple type is 25, the conservative complex type is 37, and the exaggerated complex type is 44. Combining the number of classification results and the number of correct classifications for each type of academy drawing, the classification accuracy of each type can be obtained. Regarding the classification accuracy of military school student paintings, the conservative simple type is 96.55%, the exaggerated simple type is 80.64%, the conservative complex type is 80.43%, and the exaggerated complex type is 80.00%. The classification model used in this study has the highest classification accuracy for participating paintings, with conservative simple models having the highest accuracy and exaggerated complex models having the lowest accuracy. This is because the composition and color of the latter are the most complex. The recall test results of the painting image style classification model are shown in Fig. 7.

Fig. 7. Recall rate results.

../../Resources/ieie/IEIESPC.2025.14.1.11/image7.png

In terms of the number of correctly classified pictures in Fig. 7, there are 51 for the conservative simple type, 25 for the exaggerated simple type, 37 for the conservative complex type, and 44 for the exaggerated complex type. The number of correct classifications is the highest for the conservative simple type and the lowest for the exaggerated simple type. In terms of the number of misclassifications, the conservative simple type is 11 times, the exaggerated simple type is 9 times, the conservative complex type is 6 times, and the exaggerated complex type is 9 times. Conservative and simple participants have the highest number of misclassifications in their paintings, while conservative and complex participants have the lowest number. The recall status of the model can be determined by combining the number of correct classifications and the number of incorrect classifications for each category of the model. In terms of the recall rate of the participant chart, the conservative simple type was 85.35%, the exaggerated simple type was 64.00%, the conservative complex type was 83.78%, and the exaggerated complex type was 79.54%. The painting image style classification model in this study has the highest recall rate for conservative and simple types of paintings, and the lowest recall rate for exaggerated and simple types. This is mainly because the composition of exaggerated and minimalist participants' paintings is relatively simple, and the composition information features are not obvious, which is easy to be misclassified.

4.2 Analysis of drawing description generation effect

In the analysis of the effect of drawing description generation, this study primarily compares the same type of models to test the effect. This study compared single-layer attention mechanism, double-layer attention mechanism, and simple LSTM model with the proposed three-layer attention mechanism model. The purpose is to test the accuracy and recall changes of the model under changes in sample size, that is, to determine whether the description generated by the model can accurately describe the painting. The model accuracy results are shown in Fig. 8.

Fig. 8. Accuracy change trend.

../../Resources/ieie/IEIESPC.2025.14.1.11/image8.png

In Fig. 8, the accuracy line of the single-layer attention mechanism model shows a significant decreasing trend from 85.17% to 34.23% as the sample size varies from 1 to 100. The accuracy line of the pure LSTM model also shows a decreasing trend with the increasing sample size from 90.23% to 66.18%. Compared with the single-layer attention mechanism model and the pure LSTM model, the accuracy fold of the two-layer attention mechanism model does not show a significant decreasing trend, but still decreases from 93.24 to 78.63%. The accuracy line of the three-layer attention mechanism painting description generation model also shows a decreasing trend, but only from 97.74% to 84.45%. The highest accuracy point of the drawing description generation model of the three-layer attention mechanism model is the peak in the comparison model. The lowest point is still the peak in the comparative model. The overall folding line accuracy of newspapers is above 80%. The description generated by the three-layer attention mechanism model painting description generation model is more accurate. The variation of the recall rate is shown in Fig. 9.

In Fig. 9, the recall folds of all four models show a decreasing trend as the sample size varies from 1 to 100. The downward trend of recall folding in the single-layer attention mechanism model is most significant compared to the simple LSTM model and the double-layer attention mechanism model. But as the sample size increases, when recall folding drops below 80%, only the three-layer attention mechanism model remains above 80%. The recall rate of the three-layer attention mechanism model is the best among the comparison models, fluctuating between 83.71% and 98.27%. This shows that the three-layer attention mechanism model drawing description generation model has more performance advantages in drawing description generation.

Fig. 9. Change trend of recall rate.

../../Resources/ieie/IEIESPC.2025.14.1.11/image9.png

5. Conclusion

To solve the learning difficulties caused by the lack of precise guidance in art self-learning, this study designed an interactive art teaching system for students and student to student based on the iOS system as the main platform. Among them, the interaction between the system and the trainees is achieved through a painting image style classification model. The interaction between trainees and trainees is completed through drawing descriptions to generate models. The results showed that the classification accuracy of the designed painting image style classification model was 96.55%, 80.64%, and 80.00% for the four painting types, respectively. Also, the recall rates for the four types were 85.35%, 64.00%, 83.78%, and 79.54%, respectively. It proved the effectiveness of the designed painting image style classification model. Meanwhile, the accuracy of the drawing description generation model also showed a downward trend, but only decreased from 97.74% to 84.45%. The recall rate had always remained above 80% and fluctuated between 83.71% and 98.27%. This was in an advantageous position among similar models. These data showed that the interactive art teaching system designed in the study can effectively aid beginners.

REFERENCES

1 
M. P. Grosz, J. M. Lemp, B. Rammstedt, M. Clemens, ``Lechner. Personality changes through arts education: A review and call for further research,'' Perspectives on Psychological Science, vol. 17, no. 2, pp. 360-384, 2022.DOI
2 
T. Groenendijk, A. K\'{a}rp\'{a}ti, and F. Haanstra, ``Self-assessment in art education through a visual rubric,'' International Journal of Art & Design Education, vol. 39, no. 1, pp. 153-175, 2020.DOI
3 
K. Elpus, ``Access to arts education in America: The availability of visual art, music, dance, and theater courses in US high schools,'' Arts Education Policy Review, vol. 123, no. 2, pp. 50-69, 2022.DOI
4 
K. Cosier, ``What can art and art education do in the perilous present?'' Studies in Art Education, vol. 60, no. 3, pp. 260-268, 2019.DOI
5 
R. M. Simamora, ``The challenges of online learning during the COVID-19 pandemic: An essay analysis of performing arts education students,'' Studies in Learning and Teaching, vol. 1, no. 2, pp. 86-103, 2020.DOI
6 
C. M. Thompson, ``Beyond the single story of childhood: Recognizing childism in art education practice,'' Visual Arts with Young Children, Routledge, Routledge, pp. 159-168, 2021.URL
7 
J. Ohme, T. Araujo, C. H. de Vreese, ``Mobile data donations: assessing self-report accuracy and sample biases with the iOS Screen Time function,'' Mobile Media & Communication, vol. 9, no. 2, pp. 293-313, 2021.DOI
8 
C. Zhang, W. Yi, Y. Liu, and L. Y. Song, ``STAR-IOS aided NOMA networks: Channel model approximation and performance analysis,'' IEEE Transactions on Wireless Communications, vol. 21, no. 9, pp. 6861-6876, 2022.DOI
9 
M. Chabbi, J. Lin, and R. Barik, ``An experience with code-size optimization for production iOS mobile applications,'' Proc. of IEEE/ACM International Symposium on Code Generation and Optimization (CGO), IEEE, pp. 363-377, 2021.DOI
10 
T. Groenendijk, A. K\'{a}rp\'{a}ti, and F. Haanstra, ``Self-assessment in art education through a visual rubric,'' International Journal of Art & Design Education, vol. 39, no. 1, pp. 153-175, 2020.DOI
11 
R. M. Simamora, ``The challenges of online learning during the COVID-19 pandemic: an essay analysis of performing arts education students,'' Studies in Learning and Teaching, vol. 1, no. 2, pp. 86-103, 2020.DOI
12 
Y. Long, ``Research on art innovation teaching platform based on data mining algorithm,'' Cluster Computing, vol. 22, no. 6, pp. 14943-14949, 2019.DOI
13 
Y. Gao, ``Blended teaching strategies for art design major courses in colleges,'' International Journal of Emerging Technologies in Learning (iJET), vol. 15, no. 24, pp 145-158, 2020.DOI
14 
R. Miller, ``Integrating the arts and creativity in STEM education: Emerging talent using STEAM,'' STEM Education for High-Ability Learners, Routledge, pp. 207-223, 2021.URL
15 
L. Zhang, H. Zhang, and E. Xu, ``Information entropy and elasticity analysis of the land use structure change influencing eco-environmental quality in Qinghai-Tibet Plateau from 1990 to 2015,'' Environmental Science and Pollution Research, vol. 29, no. 13, pp. 18348-18364, 2022.DOI
16 
W. Wu, J. Hou, Z. Zhang, F. F. Li, R. Zhang, L. Gao, and H. Ni, ``Information entropy-based strategy for the quantitative evaluation of extensive hyperspectral images to better unveil spatial heterogeneity in mass spectrometry imaging,'' Analytical Chemistry, vol. 94, no. 29, pp. 10355-10366, 2022.DOI
17 
H. Fukui, T. Hirakawa, and T. Yamashita, ``Attention branch network: Learning of attention mechanism for visual explanation,'' Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10705-10714, 2019.DOI
18 
X. Zhu, D. Cheng, and Z. Zhang, ``An empirical study of spatial attention mechanisms in deep networks,'' Proceedings of the IEEE/ CVF International Conference on Computer Vision, pp. 6688-6697, 2019.DOI
19 
Z. Xiong, D. Wang, X. Liu, F. S. Zhong, and X. Z. Wan, ``Pushing the boundaries of molecular representation for drug discovery with the graph attention mechanism,'' Journal of Medicinal Chemistry, vol. 63, no. 16, pp. 8749-8760, 2019.DOI
20 
S. Kara, ``An investigation of technological pedagogical and content knowledge (TPACK) competencies of pre-service visual arts teachers,'' International Journal of Technology in Education (IJTE), vol. 4, no. 3, pp. 527-541, 2021.DOI
21 
M. C. Manifold, ``The M135 e'telier: Addressing the needs of nonart majors in secondary and postsecondary art education programs,'' Studies in Art Education, vol. 60, no. 4, pp. 287-302, 2019.DOI
22 
E. Winner, L. Hetland, and S. Veenema, ``Studio thinking: How visual arts teaching can promote disciplined habits of mind,'' New directions in Aesthetics, Creativity, and the Arts, Routledge, pp. 189-206, 2020.DOI
23 
M. Chabbi, J. Lin, an dR. Barik, ``An experience with code-size optimization for production iOS mobile application,'' Proc. of IEEE/ACM International Symposium on Code Generation and Optimization (CGO), IEEE, pp. 363-377, 2021.DOI
24 
L. S\"{a}vendahl, M. Polak, P. Backeljauw, et al, ``Treatment of children with GH in the United States and Europe: Long-term follow-up from NordiNet$^{\text{\textregistered}}$ IOS and ANSWER program,'' The Journal of Clinical Endocrinology & Metabolism, vol. 104, no. 10, pp. 4730-4742, 2019.DOI
25 
L. S\"{a}vendahl, M. Polak, and P. Backeljauw, ``Long-term safety of growth hormone treatment in childhood: two large observational studies: NordiNet IOS and ANSWER,'' The Journal of Clinical Endocrinology & Metabolism, vol. 106, no. 6, pp. 1728-1741, 2021.DOI
Peng Xiao
../../Resources/ieie/IEIESPC.2025.14.1.11/author1.png

Peng Xiao was born in Hubei, China in 1978. He holds a master's degree and is an associate professor and a visiting scholar at Wuhan University. He graduated from Jingdezhen Ceramic Institute in 2001 with a bachelor's degree in arts and obtained a master's degree in management from Wuhan University in 2006. He is currently teaching at the School of Art, Wuhan Business University. His research interests include the theoretical study of ancient village architectural environments and the practical application of painting representation, as well as the theory and practice of aesthetic education in universities. In recent years, he has published a solo monograph titled ``Painting the Huizhou-style Residential Buildings'' and edited the university textbook "Architectural Sketching". He has presided over or significantly contributed to four research projects, published four pieces of artwork in CSSCI journals, and applied for three patents.