Motion design has become a crucial component of human-computer interaction (HCI), bridging the gap between interface aesthetics and functional usability. This paper provides a comprehensive review of motion design within the mobile interface, tracing its evolution from foundational animation principles to contemporary AI-augmented practices. Utilizing bibliometric analysis, we quantitatively map the research landscape and identify emerging trends within the field. We investigate the multi-dimensional theoretical framework rooted in perception, cognitive psychology, and ergonomic HCI theories to analyze the relation between motion design and user behavior. Furthermore, this study offers a systematic taxonomy of motion effects across diverse hardware platforms and investigates the workflow for AI-driven motion design generation. By synthesizing theoretical insights with practical implementation strategies, this research contributes to a robust understanding of interface dynamics and provides a strategic roadmap for the intelligent, AI-driven future of mobile user interfaces.

EEG offers a neuro-cognitive perspective to complement traditional subjective assessments in design aesthetics across digital, product, and spatial domains. This bibliometric study examines 668 publications (2006-2026) to track the evolution of EEG-based cognitive evaluation in design aesthetics. We analyzed publication trends, collaboration networks, and research frontiers, supplemented by a review of empirical datasets and methodologies. Findings reveal rapid growth since 2015, with China leading global output. Key themes have shifted from basic feasibility to aesthetic quantification, neural mechanisms, and multimodal intelligent evaluation. Current trends emphasize scenario-based applications and computationally integrated design research, reflecting a move toward more quantitative and diverse evaluation frameworks.

Material selection in computer-aided design assemblies affects product performance, manufacturability, cost, and sustainability. Existing graph-based material recommendation methods can use assembly relations, but their repeated message passing increases computational cost and may blur component-level cues when geometrically similar parts have different material labels. This paper presents a lightweight feature-embedding framework for node-level material prediction in CAD assemblies. The model projects semantic, geometric, and physical component descriptors into an expanded embedding space, concatenates the learned embedding with the original descriptors, and uses a zero-initialized adaptive residual branch to control low-level feature supplementation during training. The design uses established projection, concatenation, residual, and gating operations, and its contribution is their task-specific integration into a compact CAD material recommendation pipeline. Experiments on the Fusion 360 CAD assembly dataset show improved Micro-F1 under multiple material-tier settings with computational cost close to an MLP-based predictor.

Calligraphy is a resplendent cultural heritage, embodying a stylised visual language that conveys nuanced emotions and reflects the spirit of its creators. In this research, we aim to harness intelligent technology to analyse the emotional content and artistic styles within Chinese calligraphic works. Our dataset comprises over 5,000 calligraphy images, authored by 72 eminent Chinese calligraphers. We hypothesise that visual features can serve as key indicators for developing classification models of calligraphic emotion and styles. Given the impressive performance of machine learning techniques in image analysis, we have evaluated several algorithms, to determine the most effective model for assessing the emotional tone and stylistic elements of calligraphy. Our findings indicate that utilising ResNet18, achieves a 99.0% accuracy rate in both calligraphy emotion recognition and font identification, and Swin Transformer model outperforms others in discerning the unique writing styles of individual calligraphers. Building on these insights, we have developed a system based on the optimal model and designed to explore the rich potential of calligraphic art. This study not only deepens our appreciation for the emotional spectrum and stylistic expressions within calligraphy but also lays the groundwork for interdisciplinary studies that bridge the domains of art, culture, and technology.

Although incomplete routing tables caused by design variations, sensor faults, and record keeping lapses continue to disrupt production scheduling, most existing research still assumes data completeness. We propose a Focus-Driven Attention GAN to reconstruct missing processing times while strictly honouring precedence constraints and machine compatibility. A context preserving embedding method encodes each operation along with its neighbours, while a binary Mark-Hint strategy directs model attention explicitly towards missing values. Within its adversarial structure, a novel focus-driven discriminator provides location-specific feedback, enabling fine-grained gradient propagation. The precedence aware generator integrates forward only intra-sequence attention, cross-token attention, and machine-wise attention to effectively capture long-range data sparse dependencies. Extensive evaluations on standard job shop benchmarks and a real world engine assembly line show that our method consistently achieves superior reconstruction accuracy compared to contemporary generative imputation models, enhancing downstream schedule feasibility.

The harmony between wood grain textures and furniture design styles is a crucial determinant of furniture visual aesthetics. However, traditional methods rely on designers' subjective experience, lacking user preference data support. Moreover, there is a scarcity of related research. This study proposes a user emotional imagery-driven smart matching framework for wood grain textures with furniture styles (WTFS-SM), which optimises the design process by integrating generative AI. The steps of the framework are as follows: (1) Construct a dataset of wood grain texture imagery dimensions by combining expert interviews and user ratings, and train a CNN-based prediction model for emotional imagery mapping. (2) Extract typical furniture style feature samples through K-means clustering. (3) Generate matching schemes between wood grain textures and furniture styles using a diffusion model. (4) Utilise an interactive genetic algorithm (IGA) to integrate expert ratings and select optimal solutions. Experiments indicate that the WTFS-SM framework significantly enhances the aesthetic appeal of design proposals, operational efficiency, and user satisfaction. This study confirms that frameworks combining user emotional imagery with generative AI can provide data-driven decision support for furniture design.

In the realm of environmental protection, sustainable design has emerged as a pivotal concern. As an important aspect of sustainable design, the application of clean energy technology is crucial for product innovation. This study focuses on the innovation of zero-carbon torch design based on a clean energy solution. We explored the potential of using ammonia as an alternative to traditional fossil fuel for torch energy supply. By thoroughly analyzing the chemical properties of ammonia combustion and its potential in sustainable development, a handheld zero-carbon torch has been created. The torch design study was held to optimize combustion efficiency and reduce environmental impact. Experimental simulations were conducted to validate the design’s feasibility, confirming that the torch maintains stable combustion across diverse climatic conditions. Our structural design and energy strategy offer a viable technical solution that combines clean energy with product innovation, catering to industry needs.

This study delves into the nuanced interplay between the motion features of hand-drawn lines and their capacity to convey emotions, a relatively underexplored facet within the realm of human-computer interaction and visual art. By initiating an original experimental design, we generated a pioneering dataset, capturing both static and motion features of lines drawn to express a spectrum of emotions. Through meticulous analysis employing multivariate ordered logistic regression, we unearthed significant motion features that significantly influence emotional expression, alongside corroborating the relevance of certain static features. Our investigation extends beyond mere feature identification, exploring how these attributes correlate with emotional perceptions across a broad emotional spectrum. This research not only bridges a gap in existing literature but also lays foundational insights for future explorations into the emotional dimensions of visual art and design, offering new perspectives for enhancing creative processes and understanding the art-emotion nexus.

Calligraphy expresses people's inner mood and reflects their self-cultivation. Style and emotion plays a key role in calligraphy artistry. Consequently, we applied intelligent technology to analyze the emotion and artistic style recognition of Chinese calligraphy artworks. A dataset of 5,000 calligraphy images was established, including art works over seventy renowned Chinese calligraphists. We compared several deep-learning methods to build the best model. The results show that the optimal model achieves an accuracy of 0.9319 with MobileNet in calligraphy emotion recognition, while it achieves the best accuracy of 0.972 in calligraphy font style recognition, and the highest classification accuracy of 0.943 in calligraphy artist's style recognition respectively using ResNet50. Finally, an intelligent Chinese calligraphy emotion and style exploration application are developed based on the optimal model.

Art education is important to the overall development of children, and art training can start from improving children’s art appreciation and perception. In this study, we built an interactive painting synthesis system for children’s art education to let them know about world famous paintings and art history. The interactive learning systems can provide engaging ways of art education based on image synthesis, style transfer and animation generation technology. Specifically, an image dataset of famous paintings was established for painting synthesis and an application “ARTIST” was developed based on First Order Motion Method. The system “ARTIST” can attract children’s active participation via interesting art creation interaction and enhance their learning motivation in their exploration.

The prevalence of stroke continues to increase with the global aging. Based on the motor imagery (MI) brain–computer interface (BCI) paradigm and virtual reality (VR) technology, we designed and developed an upper-limb rehabilitation exoskeleton system (VR-ULE) in the VR scenes for stroke patients. The VR-ULE system makes use of the MI electroencephalogram (EEG) recognition model with a convolutional neural network and squeeze-and-excitation (SE) blocks to obtain the patient’s motion intentions and control the exoskeleton to move during rehabilitation training movement. Due to the individual differences in EEG, the frequency bands with optimal MI EEG features for each patient are different. Therefore, the weight of different feature channels is learned by combining SE blocks to emphasize the useful information frequency band features. The MI cues in the VR-based virtual scenes can improve the interhemispheric balance and the neuroplasticity of patients. It also makes up for the disadvantages of the current MI-BCIs, such as single usage scenarios, poor individual adaptability, and many interfering factors. We designed the offline training experiment to evaluate the feasibility of the EEG recognition strategy, and designed the online control experiment to verify the effectiveness of the VR-ULE system. The results showed that the MI classification method with MI cues in the VR scenes improved the accuracy of MI classification (86.49% ± 3.02%); all subjects performed two types of rehabilitation training tasks under their own models trained in the offline training experiment, with the highest average completion rates of 86.82% ± 4.66% and 88.48% ± 5.84%. The VR-ULE system can efficiently help stroke patients with hemiplegia complete upper-limb rehabilitation training tasks, and provide the new methods and strategies for BCI-based rehabilitation devices.

Prolonged use of a keyboard or mouse can lead to continuous bending and twisting of the wrist, which can lead to muscle fatigue and pain. Finger exercises can reduce the hand muscle fatigue effectively. However, the repetitive exercises will bring boring and tedious experience. In this paper, we developed a hand rehabilitation exercise system based on gesture recognition technology and gamification design, which provides a more enjoyable, personalized and effective rehabilitation experience. The hand fatigue relief system was designed with the fusion of game elements and gesture interaction, which can enhance users’ engagement and their rehabilitation motivation.

Emotion is an essential aspect that influences the learning outcome and MOOCs learning experience. This research investigates the possibility of employing the fusion characteristics of video and users' eye tracking features to assess MOOC learners' emotions. In the experiment, we gathered eye tracking data from MOOCs students. We applied OpenSmile and OpenCV to extract the video features. And a fusion data set with the eye-tracking data and video features was established in the experiment. Machine learning method is utilized to explore the optimal model using fused feature data sets, and the impacts of multimodal features on emotion recognition are discussed. Our findings demonstrate that the fusion feature can achieve an accuracy rate of 86.3%, which outperformed the single-modal feature sets. According to the study, this technique can be used to learn students' emotion status in their MOOC study and predict the students' affective perception on MOOCs videos.

In view of the importance of neck strength training and the lack of adequate training equipment, this study designed a new oscillating hydraulic trainer (OHT) of neck based on oscillating hydraulic damper. We used surface electromyography (sEMG) and subjective ratings to evaluate the neck OHT and compared the results with a simple hat trainer (HATT) and traditional weight trainer (TWT) to verify the feasibility and validity of the OHT. Under similar exercise conditions, 12 subjects performed a set of neck flexion and extension exercise with these 3 trainers. The sEMG signals of targeted muscles were collected in real time, and subjects were asked to complete subjective evaluations of product usability after exercise. The results showed that the root mean square (RMS%) of sEMG indicated that the OHT could provide two-way resistance and train the flexors and extensors simultaneously. The overall degree of muscle activation with OHT was higher than that with the other two trainers in one movement cycle. In terms of resistance characteristics exhibited by the sEMG waveform, duration (D) with OHT was significantly longer than HATT and TWT when exercising at a high speed, while Peak Timing (PT) was later. The ratings of product usability and performing usability of OHT were remarkably higher than that of HATT and TWT. Based on the above results, the OHT was proved to be more suitable for strength training, such as neck muscles, which were getting more attention gradually, but lacked mature and special training equipment.

Third-person pain is an interesting empathy phenomenon that human has the ability to infer characters of other sufferers’ pain by observing their behavior. In the literature, existing studies suggesting that first-person and third-person pain share common features of neuroimage, which indicates that pain behavior will cause influence on both sufferer and the observer. Consequently, it is significant to explore the third-person effects upon the observer. In this study, the evaluation and recognition of third-person pain experience was studied based on user physiological signal analysis. We built a third-person pain multimodal physiological features dataset and applied machine learning methods to explore a third-person pain experience recognition model. A classification accuracy of 95.83% was obtained in third-person pain degree recognition, which demonstrates the effectiveness of our approach. The proposed study shed light on the guiding future exploration of determinants of third-person pain process and empathy intelligence.

Much research has been done on the relationship between emotions and colors, and many color scheme recommendation tools have been developed. These tools can often help recommend a suitable color scheme to express a certain emotion, but it is still difficult to design satisfying schemes based on such recommendations. In this study, based on the Color Scheme Bible, Compact Edition, we developed Emocolor, which allows professionals to generate a large number of color schemes based on emotional words or emotional images. It is also used to iteratively optimize the generated color schemes based on an interactive genetic algorithm to find the color scheme design that best matches the user's emotion. Through the evaluation of tram color schemes based on emotional words, Emocolor can effectively generate tram color schemes that match users' emotions. It can also effectively transfer the emotion of an image with a single dominant color to a tram color scheme. In short, Emocolor can help professionals describe the relationship between emotion and color in a more accurate way and design emotional color schemes. Emocolor can be applied in advertising design, product design, interior design, and other fields.

Users' viewing behavior could affect their perception and evaluation of design works. Taking into account users' visual attention as a subjective cognition cue, we used eye-tracking evidence to identify users' focus areas for further analysis. We conducted experiments to extract the image features of design images and the reviewers' eye-tracking data, aiming to predict the product design ranking in the competition through fusion data analysis. In particular, we collected 1,504 product design images from a design competition. Four deep convolutional neural networks were selected to explore the best aesthetics computation model. The experimental results show that using design images and eye-tracking data fusion can improve the model prediction performance. Finally, MobileNet-V3 achieves the highest classification accuracy of 74.75%. This suggests the proposed method can provide useful insights into personalized aesthetics evaluation and user-centered design perception.

Visual art works contain a lot of tacit knowledge that is difficult to accurately express in words. If they are expressed quantitatively in a computable form, it helps to apply this part of tacit knowledge to a wider field. Chinese calligraphy style carriers have tacit knowledge of Chinese cultural characteristics, which our research quantitatively interprets. In our study, 33 interpretable features were designed and summarized, and the random forest classification was adopted. As a result, we found that only 8 computational features with concise mathematic form were needed to interpret the differences between the five writing styles of Chinese calligraphy with an accuracy of 66.7 %. Based on these features and the evaluation of five calligraphy styles, we find that some combination of features can cause people’s perception of a particular style and establish the relationship between objective features and people’s subjective feelings. The results can provide inspiration for the creation of artists and designers, and have potential applications in the fields of psychology, design, and human-computer interaction.

Traffic congestion can lead to negative driving emotions, significantly increasing the likelihood of traffic accidents. Reducing negative driving emotions as a means to mitigate speeding, reckless overtaking, and aggressive driving behaviors is a viable approach. Among the potential methods, affective speech has been considered one of the most promising. However, research on humor-based affective speech interventions in the context of driving negative emotions is scarce, and the utilization of electroencephalogram (EEG) signals for emotion detection in humorous audio studies remains largely unexplored. Therefore, our study first designed a highly realistic experiment scenario to induce negative emotions experienced by drivers in congested traffic conditions. Subsequently, we collected drivers’ EEG signals and subjective questionnaire ratings during the driving process. By employing one-way analysis of variance (ANOVA) and t tests, we analyzed the data to validate the success of our experiment in inducing negative emotions in drivers during congested road conditions and to assess the effectiveness of humorous audio in regulating drivers’ negative emotions. The results indicated that humorous audio effectively alleviated drivers’ negative emotions in congested road conditions, with a 145.84% increase in arousal and a 93.55% increase in valence ratings compared to control conditions. However, it should be noted that humorous audio only restored drivers’ emotions to the level experienced during normal driving. Our findings offer novel insights into regulating drivers’ negative emotions during congested road conditions.

Leveraging the power of computational methods, AI can perform effective strategies in intelligent design. Researchers are pushing the boundaries of AI, developing computational systems to solve complex questions. The authors investigate the association of user preference for UI and deep image features, aiming to predict user preference level using deep convolutional neural networks (DCNNs) trained on a UI design image dataset. A total of 12,186 UI design images were collected from UI.cn and DOOOOR.com. Users' views and likes can help understand the implicit user preference level, which is set as the ground-truth annotation for the dataset. Six DCNNs, including VGG-19, InceptionNet-V3, MobileNet, EfficientNet, ResNet-50 and NASNetLarge were trained to learn the user preference of UI images. The experiment achieves an optimal result with a mean-squared error of 0.000214 and a mean absolute error of 0.0103 based on EfficientNet, which indicates that the proposed method provides the possibility in learning the pattern of user aesthetics preference for UI design. On the basis of the prediction model, a mobile application named 'HotUI' was developed for UI design recommendations.