Minkai Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Minkai Sun has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
Yanyan Bao
School of Architecture and Urban Planning, Suzhou University of Science and Technology; 215000, 1701, Binhelu, Suzhou, China
3 shared publications- Akiyoshi Honden
Department of Agro-Environmental Sciences, Faculty of Agriculture, Kyushu University, Fukuoka 8190395, Japan
1 shared publication - Chang Li
School of Architecture and Urban Planning, Suzhou University of Science and Technology, Suzhou 215009, China
1 shared publication - Changan Sun
School of Education, Suzhou University of Science and Technology, Suzhou 215000, China
1 shared publication Jian Zhang
1 shared publication- Koichiro Ohnuki
Department of Biological and Environmental Chemistry, Kindai University, 11-6 Kayanomori, Iizuka 820-8555, Japan
1 shared publication
Latest Publications
- Open Access
- 20 May 2026
- Published by Frontiers Media SA in Frontiers in Built Environment
- Vol. 12, 1825552
Research on AI-assisted generation of jiangnan classical garden architectural scenes based on LLM and LDM
Classical garden visualization commonly suffers from high production cost, time-consuming production processes, and low efficiency in iterative design. To address these limitations, this study proposes a domain-oriented AI-assisted design framework for Jiangnan classical garden architectural scenes based on a large language model (LLM) and a latent diffusion model (LDM). Rather than emphasizing the generic combination of LLMs and diffusion models, the study focuses on constructing a workflow that links natural-language parsing, structured semantic representation, distributed LoRA-based coordinated generation, and multi-level evaluation within a strongly domain-constrained design context. Specifically, the LLM is fine-tuned through supervised fine-tuning using classical garden literature and standardized samples to generate structured prompts aligned with professional design semantics, while a distributed hierarchical LoRA strategy is applied to the LDM to separate architectural morphology control from environment-level scene expression. A trigger-word-based coordination mechanism further supports rule-driven invocation and dynamic weighting of architectural and environmental LoRA models. Through this framework, the study seeks to address the unstable architectural type recognition, insufficient spatial-semantic mapping, and stylistic deviation often observed when general-purpose generation models are applied to traditional garden architectural scenes. Evaluation results indicate that more than 75% of the generated images reach a professionally usable level in spatial organization and artistic-conception restoration, while over 80% exhibit traditional garden aesthetics and visual authenticity. The proposed framework improves generation efficiency and provides a domain-oriented technical pathway for controllable generation, design iteration, design communication, and literature-informed reconstruction of Jiangnan classical garden scenes.
- 20 May 2026
- Published by Frontiers Media SA in Frontiers in Built Environment
- Vol. 12, 1825552
Open AccessResearch on AI-assisted generation of jiangnan classical garden architectural scenes based on LLM and LDM
Classical garden visualization commonly suffers from high production cost, time-consuming production processes, and low efficiency in iterative design. To address these limitations, this study proposes a domain-oriented AI-assisted design framework for Jiangnan classical garden architectural scenes based on a large language model (LLM) and a latent diffusion model (LDM). Rather than emphasizing the generic combination of LLMs and diffusion models, the study focuses on constructing a workflow that links natural-language parsing, structured semantic representation, distributed LoRA-based coordinated generation, and multi-level evaluation within a strongly domain-constrained design context. Specifically, the LLM is fine-tuned through supervised fine-tuning using classical garden literature and standardized samples to generate structured prompts aligned with professional design semantics, while a distributed hierarchical LoRA strategy is applied to the LDM to separate architectural morphology control from environment-level scene expression. A trigger-word-based coordination mechanism further supports rule-driven invocation and dynamic weighting of architectural and environmental LoRA models. Through this framework, the study seeks to address the unstable architectural type recognition, insufficient spatial-semantic mapping, and stylistic deviation often observed when general-purpose generation models are applied to traditional garden architectural scenes. Evaluation results indicate that more than 75% of the generated images reach a professionally usable level in spatial organization and artistic-conception restoration, while over 80% exhibit traditional garden aesthetics and visual authenticity. The proposed framework improves generation efficiency and provides a domain-oriented technical pathway for controllable generation, design iteration, design communication, and literature-informed reconstruction of Jiangnan classical garden scenes.
- Open Access
- 30 July 2025
- Published by MDPI AG in Architecture
- Vol. 5 (3) , 57
Exploring the Garden Design and Underlying Philosophy of Lion Grove as a Chan Garden During the Yuan Dynasty
Lion Grove was established in 1342 during the Yuan Dynasty and is one of the four most famous classical gardens in China. It was recognized as a UNESCO World Heritage Site in 2000. Although Lion Grove is now regarded as a private garden representing the culture of Confucian scholars, it was originally a Chan Buddhist garden during its inception in the Yuan Dynasty. This study examines the natural landscapes of Lion Grove at its inception, focusing on four main aspects: the philosophy of nature, planning intent, creators, and the philosophical ideas they represent. Key findings include the following: Firstly, Lion Grove’s attitude towards nature is rooted in China’s indigenous culture, making it both a physical expression of Chan philosophy and a space reflecting the scholar–bureaucrats’ vision of an ideal landscape. Secondly, from the perspective of landscape planning, the Lion Grove of the Yuan Dynasty placed greater emphasis on natural elements compared to its modern counterpart, with rock landscapes serving as the core element throughout the garden. Thirdly, hermitic philosophy emerged as a significant cultural theme alongside Chan Buddhism during the Yuan Dynasty. Fourthly, the landscape elements of Lion Grove symbolize Chan Buddhist wisdom and the hermit’s idealism, with poetry playing a key role in conveying these cultural ideals, preserving the site’s early philosophical significance.
- 30 July 2025
- Published by MDPI AG in Architecture
- Vol. 5 (3) , 57
Open AccessExploring the Garden Design and Underlying Philosophy of Lion Grove as a Chan Garden During the Yuan Dynasty
Lion Grove was established in 1342 during the Yuan Dynasty and is one of the four most famous classical gardens in China. It was recognized as a UNESCO World Heritage Site in 2000. Although Lion Grove is now regarded as a private garden representing the culture of Confucian scholars, it was originally a Chan Buddhist garden during its inception in the Yuan Dynasty. This study examines the natural landscapes of Lion Grove at its inception, focusing on four main aspects: the philosophy of nature, planning intent, creators, and the philosophical ideas they represent. Key findings include the following: Firstly, Lion Grove’s attitude towards nature is rooted in China’s indigenous culture, making it both a physical expression of Chan philosophy and a space reflecting the scholar–bureaucrats’ vision of an ideal landscape. Secondly, from the perspective of landscape planning, the Lion Grove of the Yuan Dynasty placed greater emphasis on natural elements compared to its modern counterpart, with rock landscapes serving as the core element throughout the garden. Thirdly, hermitic philosophy emerged as a significant cultural theme alongside Chan Buddhism during the Yuan Dynasty. Fourthly, the landscape elements of Lion Grove symbolize Chan Buddhist wisdom and the hermit’s idealism, with poetry playing a key role in conveying these cultural ideals, preserving the site’s early philosophical significance.
- Top Cited PapersOpen Access
- 22 February 2025
- Published by Springer Nature in npj Heritage Science
- Vol. 13 (1) , 1-11
Effects of reflecting water on viewers’ visual attention, spatial perception, and emotional responses: a case study in Chinese heritage garden
- 22 February 2025
- Published by Springer Nature in npj Heritage Science
- Vol. 13 (1) , 1-11
Top Cited PapersOpen AccessEffects of reflecting water on viewers’ visual attention, spatial perception, and emotional responses: a case study in Chinese heritage garden
- Open Access
- 15 May 2024
- Published by Springer Nature in Research Square
Effects of Reflecting Water on Viewers’ Visual Perception and Subjective Evaluation: A Case Study in Classic Chinese Garden
Water is considered one of the most important elements in landscape design. Studies claim that water expands perceived spatial dimension. In classic Chinese gardens, water is usually used to expand limited garden spaces. However, this information is mostly experience-based. Therefore, this exploratory study compared the differences in characteristics of the observer's visual and cognitive processing using a typical classic Chinese garden scenery picture with and without water through experiment (repeated measures design). A total of 64 participants were asked to randomly view two pictures respectively. The results showed that water reflection attracts more visual attention than pavement. Moreover, reflections can reduce the proportion of visual attention of a reflected substance and lower its visual attractiveness. The estimation of the actual area size showed that water was not effective for people’s spatial perception. However, participants felt a greater sense of spaciousness when viewing scenes that included water. These results suggest that reflecting on the water enhanced people’s appreciation of the scene.
- 15 May 2024
- Published by Springer Nature in Research Square
Open AccessEffects of Reflecting Water on Viewers’ Visual Perception and Subjective Evaluation: A Case Study in Classic Chinese Garden
Water is considered one of the most important elements in landscape design. Studies claim that water expands perceived spatial dimension. In classic Chinese gardens, water is usually used to expand limited garden spaces. However, this information is mostly experience-based. Therefore, this exploratory study compared the differences in characteristics of the observer's visual and cognitive processing using a typical classic Chinese garden scenery picture with and without water through experiment (repeated measures design). A total of 64 participants were asked to randomly view two pictures respectively. The results showed that water reflection attracts more visual attention than pavement. Moreover, reflections can reduce the proportion of visual attention of a reflected substance and lower its visual attractiveness. The estimation of the actual area size showed that water was not effective for people’s spatial perception. However, participants felt a greater sense of spaciousness when viewing scenes that included water. These results suggest that reflecting on the water enhanced people’s appreciation of the scene.
- Open Access
- 1 April 2024
- Published by China Science Publishing & Media Ltd. in Landscape Architecture
- Vol. 31 (4) , 79-86
Research Progress on the Application of Eye Tracking Technology in Landscape Architecture
Objective Eye tracking technology was initially widely applied in fields such as medicine, education, transportation, and human-computer interaction. Subsequently, it gradually found its way into geography, tourism, and landscape research. This technology can collect and analyze behavioral data of individuals in landscape environments, revealing the inherent correlation between visual attention and spatial features. In the field of landscape architecture, eye tracking technology is primarily used for landscape evaluation and perception, yielding numerous outcomes. However, there is a lack of systematic research on the hotspots and trends of eye tracking technology in landscape architecture. This research reviews and summarizes the current application status of this technology in landscape architecture research, aiming to provide a reference for future landscape architecture research based on eye tracking technology. Method In this research, CiteSpace software is employed for bibliometric analysis and construction of visual knowledge map in an effort to figure out the development trends, disciplinary distribution characteristics, and keyword clustering of eye tracking technology. The research aims to explore the evolutionary trends, dynamics, hotspots, and frontiers of eye tracking technology in the field of landscape architecture research, thereby revealing the overall development status of relevant researches. The English literature data are sourced from the Web of Science (WoS) core collection. The retrieval uses the following query: “TS = (eye movement OR eye tracking) AND (landscape OR visual landscape preference) AND (visual attention OR perception).” The document type selected for retrieval is “article (academic journal paper)”, and the articles retrieved are mainly designed to adequately reflect the characteristics and development trends of the research topic. The retrieval of English literature was conducted on May 10, 2023, with the literature retrieved spanning the period from 2000 to 2022 and a total of 98 articles meeting the specified criteria being retrieved. Chinese literature data are obtained from the China National Knowledge Infrastructure (CNKI). The retrieval uses the keywords “eye tracking” AND “landscape” OR “landscape architecture”, focusing on article titles. The retrieval of Chinese literature was conducted on May 10, 2023, covering academic journal papers published during the period from 2000 and 2022, with a total of 105 articles meeting the specified criteria being retrieved. Results The research uses CiteSpace to integrate high-frequency keywords and keyword clustering. By combining the authority and latest literature of representative scholars within each cluster, the foreign research direction of eye tracking technology in landscape architecture is summarized as landscape perception and therapeutic landscape. Meanwhile, the domestic research focuses on eye tracking technology in landscape architecture include landscape perception, landscape evaluation, and therapeutic landscape. The research on therapeutic landscape primarily focuses on the positive impact of the environment on human stress recovery and attention fatigue improvement. In existing researches, eye tracking technology primarily records the visual attention characteristics of individuals in natural environments or urban settings dominated by natural elements, and analyzes the therapeutic effects of different environments. In landscape perception research, eye tracking technology is primarily employed to explore differences between various landscape types, such as urban or natural landscape. It is also used to investigate the impact of the quantity and design techniques of different elements within similar landscapes on landscape perception. Landscape evaluation is the outcome of the interaction between evaluators’ psychological activities and the quality of landscape, which entails evaluators to make value judgments on the external and functional aspects of landscape based on visual perception. In the future, landscape architecture research based on eye tracking technology may pay more attention to the following three aspects. 1) Expansion of the interdisciplinary integration scope of the research on eye tracking technology. Eye tracking technology can be applied in areas such as ecological perception, making it possible to evaluate the reactions of different populations to ecological landscape, identify the change trends of ecological environments, and thus optimize ecological environments. 2) Innovative research involving the combination of eye tracking technology with various other technologies. Eye tracking technology can synergize with physiological sensing technologies such as heart rate variability, magnetic resonance imaging, and functional near-infrared spectroscopy. 3) Research on the application of eye tracking technology in the health research field. Future research may combine virtual reality technology, biofeedback technology, and other approaches to obtain more comprehensive empirical data support. Conclusion This research summarizes the progress of landscape architecture research based on eye tracking technology, elucidating the key roles of eye tracking technology in the research on therapeutic landscape, landscape perception, landscape evaluation, etc. In the future, the research on landscape architecture based on eye tracking technology is expected to evolve into an interdisciplinary and technologically integrated research system.
- 1 April 2024
- Published by China Science Publishing & Media Ltd. in Landscape Architecture
- Vol. 31 (4) , 79-86
Open AccessResearch Progress on the Application of Eye Tracking Technology in Landscape Architecture
Objective Eye tracking technology was initially widely applied in fields such as medicine, education, transportation, and human-computer interaction. Subsequently, it gradually found its way into geography, tourism, and landscape research. This technology can collect and analyze behavioral data of individuals in landscape environments, revealing the inherent correlation between visual attention and spatial features. In the field of landscape architecture, eye tracking technology is primarily used for landscape evaluation and perception, yielding numerous outcomes. However, there is a lack of systematic research on the hotspots and trends of eye tracking technology in landscape architecture. This research reviews and summarizes the current application status of this technology in landscape architecture research, aiming to provide a reference for future landscape architecture research based on eye tracking technology. Method In this research, CiteSpace software is employed for bibliometric analysis and construction of visual knowledge map in an effort to figure out the development trends, disciplinary distribution characteristics, and keyword clustering of eye tracking technology. The research aims to explore the evolutionary trends, dynamics, hotspots, and frontiers of eye tracking technology in the field of landscape architecture research, thereby revealing the overall development status of relevant researches. The English literature data are sourced from the Web of Science (WoS) core collection. The retrieval uses the following query: “TS = (eye movement OR eye tracking) AND (landscape OR visual landscape preference) AND (visual attention OR perception).” The document type selected for retrieval is “article (academic journal paper)”, and the articles retrieved are mainly designed to adequately reflect the characteristics and development trends of the research topic. The retrieval of English literature was conducted on May 10, 2023, with the literature retrieved spanning the period from 2000 to 2022 and a total of 98 articles meeting the specified criteria being retrieved. Chinese literature data are obtained from the China National Knowledge Infrastructure (CNKI). The retrieval uses the keywords “eye tracking” AND “landscape” OR “landscape architecture”, focusing on article titles. The retrieval of Chinese literature was conducted on May 10, 2023, covering academic journal papers published during the period from 2000 and 2022, with a total of 105 articles meeting the specified criteria being retrieved. Results The research uses CiteSpace to integrate high-frequency keywords and keyword clustering. By combining the authority and latest literature of representative scholars within each cluster, the foreign research direction of eye tracking technology in landscape architecture is summarized as landscape perception and therapeutic landscape. Meanwhile, the domestic research focuses on eye tracking technology in landscape architecture include landscape perception, landscape evaluation, and therapeutic landscape. The research on therapeutic landscape primarily focuses on the positive impact of the environment on human stress recovery and attention fatigue improvement. In existing researches, eye tracking technology primarily records the visual attention characteristics of individuals in natural environments or urban settings dominated by natural elements, and analyzes the therapeutic effects of different environments. In landscape perception research, eye tracking technology is primarily employed to explore differences between various landscape types, such as urban or natural landscape. It is also used to investigate the impact of the quantity and design techniques of different elements within similar landscapes on landscape perception. Landscape evaluation is the outcome of the interaction between evaluators’ psychological activities and the quality of landscape, which entails evaluators to make value judgments on the external and functional aspects of landscape based on visual perception. In the future, landscape architecture research based on eye tracking technology may pay more attention to the following three aspects. 1) Expansion of the interdisciplinary integration scope of the research on eye tracking technology. Eye tracking technology can be applied in areas such as ecological perception, making it possible to evaluate the reactions of different populations to ecological landscape, identify the change trends of ecological environments, and thus optimize ecological environments. 2) Innovative research involving the combination of eye tracking technology with various other technologies. Eye tracking technology can synergize with physiological sensing technologies such as heart rate variability, magnetic resonance imaging, and functional near-infrared spectroscopy. 3) Research on the application of eye tracking technology in the health research field. Future research may combine virtual reality technology, biofeedback technology, and other approaches to obtain more comprehensive empirical data support. Conclusion This research summarizes the progress of landscape architecture research based on eye tracking technology, elucidating the key roles of eye tracking technology in the research on therapeutic landscape, landscape perception, landscape evaluation, etc. In the future, the research on landscape architecture based on eye tracking technology is expected to evolve into an interdisciplinary and technologically integrated research system.
- Open Access
- 23 March 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (7) , 3799
The Perceived Restorative Quality of Viewing Various Types of Urban and Rural Scenes: Based on Psychological and Physiological Responses
Attention restoration theory argues that the type of visual scene is important; however, related research is mostly based on a dichotomous comparison between natural and urban environments. Few studies have evaluated complex scenes comprising both natural and artificial elements. Therefore, we compared the differences between four types of environments: urban artificial scenes, urban natural scenes, rural artificial scenes, and rural natural scenes—using a survey based on the Perceived Restorativeness Scale (PRS), perception complexity scoring, and eye tracking. Participants (N = 119) viewed photographs in a random order. The results showed significant differences between the visual landscape scores and eye-tracking data for each type of visual image: PRS, perception complexity, average fixation duration, and mean pupil size. Rural natural scenes had a higher restoration effect than the other scenes. Waterscapes and well-maintained vegetation had positive correlations between the typical landscape element indices and restorative benefits in different scene types. Contrastingly, weeds and hardscapes showed negative correlations, which can be attributed to the maintenance of these typical elements. The harmony of elements with circumstances in a scene was a key factor. The results provide a reference for urban and rural landscape planning and design to improve perceived restorative quality.
- 23 March 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (7) , 3799
Open AccessThe Perceived Restorative Quality of Viewing Various Types of Urban and Rural Scenes: Based on Psychological and Physiological Responses
Attention restoration theory argues that the type of visual scene is important; however, related research is mostly based on a dichotomous comparison between natural and urban environments. Few studies have evaluated complex scenes comprising both natural and artificial elements. Therefore, we compared the differences between four types of environments: urban artificial scenes, urban natural scenes, rural artificial scenes, and rural natural scenes—using a survey based on the Perceived Restorativeness Scale (PRS), perception complexity scoring, and eye tracking. Participants (N = 119) viewed photographs in a random order. The results showed significant differences between the visual landscape scores and eye-tracking data for each type of visual image: PRS, perception complexity, average fixation duration, and mean pupil size. Rural natural scenes had a higher restoration effect than the other scenes. Waterscapes and well-maintained vegetation had positive correlations between the typical landscape element indices and restorative benefits in different scene types. Contrastingly, weeds and hardscapes showed negative correlations, which can be attributed to the maintenance of these typical elements. The harmony of elements with circumstances in a scene was a key factor. The results provide a reference for urban and rural landscape planning and design to improve perceived restorative quality.
- Open Access
- 7 February 2022
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 19 (3) , 1856
Physiological and Psychological Effects of Volatile Organic Compounds from Dried Common Rush (Juncus effusus L. var. decipiens Buchen.) on Humans
This study compared the participants’ physiological responses and subjective evaluations of air scented with different concentrations of common rush (Juncus effusus L. var. decipiens Buchen.) (30 g and 15 g, with fresh air as a control). We asked 20 participants to complete a series of visual discrimination tasks while inhaling two different air samples. We evaluated (1) brain activity, (2) autonomic nervous activity, and (3) blood pressure and pulse rate, (4) in combination with self-evaluation. In addition, we quantified the concentrations of volatile organic compounds. The participants reported the scent to be sour, pungent, and smelly; this impression was likely caused by hexanal and acetic acid. Although the self-evaluations showed that participants did not enjoy the scent, their alpha amplitudes of electroencephalogram and parasympathetic nervous activity were increased, suggesting that participants were relaxed in this atmosphere. Moreover, a lower concentration resulted in a greater induction of relaxation. While the air was not pleasant-smelling, the volatile organic compounds present had a positive psychophysiological impact.
- 7 February 2022
- Published by MDPI AG in International Journal of Environmental Research and Public Health
- Vol. 19 (3) , 1856
Open AccessPhysiological and Psychological Effects of Volatile Organic Compounds from Dried Common Rush (Juncus effusus L. var. decipiens Buchen.) on Humans
This study compared the participants’ physiological responses and subjective evaluations of air scented with different concentrations of common rush (Juncus effusus L. var. decipiens Buchen.) (30 g and 15 g, with fresh air as a control). We asked 20 participants to complete a series of visual discrimination tasks while inhaling two different air samples. We evaluated (1) brain activity, (2) autonomic nervous activity, and (3) blood pressure and pulse rate, (4) in combination with self-evaluation. In addition, we quantified the concentrations of volatile organic compounds. The participants reported the scent to be sour, pungent, and smelly; this impression was likely caused by hexanal and acetic acid. Although the self-evaluations showed that participants did not enjoy the scent, their alpha amplitudes of electroencephalogram and parasympathetic nervous activity were increased, suggesting that participants were relaxed in this atmosphere. Moreover, a lower concentration resulted in a greater induction of relaxation. While the air was not pleasant-smelling, the volatile organic compounds present had a positive psychophysiological impact.