Pub Date : 2025-12-30DOI: 10.1016/j.displa.2025.103336
Yi-Ming Li , Wen Meng , Chih-Yu Tsai , Tsung-Xian Lee
This study presents a calibration methodology designed to enhance the chromaticity and luminance accuracy of LCD monitors under low-luminance conditions, specifically targeting cost-effective medical display applications. The proposed system integrates a low-cost color sensor with a polynomial regression-based model, enhanced by automated sampling and low-luminance compensation techniques. Compared to conventional calibration workflows, the proposed system reduces the number of required samples by more than 50% while achieving comparable or superior accuracy, particularly under low-luminance conditions. This is enabled by a novel combination of luminance-aware automated sampling and perceptually guided compensation mechanisms. The automated sampling strategy significantly reduces the number of required calibration samples from 96 to 44 while maintaining high calibration accuracy, achieving an average luminance error (ΔL) of 0.606% and a color difference (ΔE) of 0.091. The low-luminance compensation algorithm mitigates accuracy degradation in darker regions, ensuring compliance with stringent medical-grade performance standards. These results demonstrate that high-precision calibration can be achieved using economical color sensors, offering a practical and scalable solution for medical-grade LCDs.
{"title":"A polynomial regression-based calibration method for enhancing chromaticity and luminance accuracy at low luminance levels of LCDs with automated sampling and compensation mechanisms","authors":"Yi-Ming Li , Wen Meng , Chih-Yu Tsai , Tsung-Xian Lee","doi":"10.1016/j.displa.2025.103336","DOIUrl":"10.1016/j.displa.2025.103336","url":null,"abstract":"<div><div>This study presents a calibration methodology designed to enhance the chromaticity and luminance accuracy of LCD monitors under low-luminance conditions, specifically targeting cost-effective medical display applications. The proposed system integrates a low-cost color sensor with a polynomial regression-based model, enhanced by automated sampling and low-luminance compensation techniques. Compared to conventional calibration workflows, the proposed system reduces the number of required samples by more than 50% while achieving comparable or superior accuracy, particularly under low-luminance conditions. This is enabled by a novel combination of luminance-aware automated sampling and perceptually guided compensation mechanisms. The automated sampling strategy significantly reduces the number of required calibration samples from 96 to 44 while maintaining high calibration accuracy, achieving an average luminance error (Δ<em>L</em>) of 0.606% and a color difference (Δ<em>E</em>) of 0.091. The low-luminance compensation algorithm mitigates accuracy degradation in darker regions, ensuring compliance with stringent medical-grade performance standards. These results demonstrate that high-precision calibration can be achieved using economical color sensors, offering a practical and scalable solution for medical-grade LCDs.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103336"},"PeriodicalIF":3.4,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1016/j.displa.2025.103331
Jin Ai , Gan Pei , Bitao Ma , Menghan Hu , Jian Zhang
Thermal imaging offers a viable approach for contactless posture monitoring due to its privacy-preserving nature and ability to capture residual thermal patterns. Existing methods exhibit limited generalization capabilities across different materials and thermal decay stages, coupled with a lack of reliable physical interpretability. To address these challenges, this study proposes an integrated paradigm combining generative data augmentation, visual transformer classification, and finite element (FE) simulation. The proposed pipeline first enhances data diversity through a generative model, then employs a Transformer-based classifier to achieve accurate recognition of 9 sitting postures. Finally, a heat conduction model is constructed to simulate the real thermal decay temperature field, decoding the influence of material and time on buttock thermal patterns. Through this paradigm, we identify a critical temperature difference threshold of 2.6 0.06 K, beyond which model performance significantly degrades. Systematic analysis demonstrates that maintaining surface temperatures above this threshold during the initial 30 s enables the model to sustain accuracy above 85%. Furthermore, we quantified the direct impact of material thermophysical parameters on the effective detection window, revealing that materials with lower thermal conductivity (e.g., plastics) extend reliable identification duration. Validation on an independent test set featuring two materials and varying decay durations demonstrated a classification accuracy of 0.9162. This study establishes a thermal imaging-based posture analysis paradigm, providing a theoretical foundation and practical solutions for real-world applications in privacy-sensitive scenarios by decoding buttock thermal patterns. The dataset and code supporting this study are publicly available at: https://github.com/AJ-1995/Thermal-Memory-of-Chair.
{"title":"Thermal memory of chair: Physics-guided generalization for sitting posture inference using thermal imaging","authors":"Jin Ai , Gan Pei , Bitao Ma , Menghan Hu , Jian Zhang","doi":"10.1016/j.displa.2025.103331","DOIUrl":"10.1016/j.displa.2025.103331","url":null,"abstract":"<div><div>Thermal imaging offers a viable approach for contactless posture monitoring due to its privacy-preserving nature and ability to capture residual thermal patterns. Existing methods exhibit limited generalization capabilities across different materials and thermal decay stages, coupled with a lack of reliable physical interpretability. To address these challenges, this study proposes an integrated paradigm combining generative data augmentation, visual transformer classification, and finite element (FE) simulation. The proposed pipeline first enhances data diversity through a generative model, then employs a Transformer-based classifier to achieve accurate recognition of 9 sitting postures. Finally, a heat conduction model is constructed to simulate the real thermal decay temperature field, decoding the influence of material and time on buttock thermal patterns. Through this paradigm, we identify a critical temperature difference threshold of 2.6 <span><math><mo>±</mo></math></span> 0.06 K, beyond which model performance significantly degrades. Systematic analysis demonstrates that maintaining surface temperatures above this threshold during the initial 30 s enables the model to sustain accuracy above 85%. Furthermore, we quantified the direct impact of material thermophysical parameters on the effective detection window, revealing that materials with lower thermal conductivity (e.g., plastics) extend reliable identification duration. Validation on an independent test set featuring two materials and varying decay durations demonstrated a classification accuracy of 0.9162. This study establishes a thermal imaging-based posture analysis paradigm, providing a theoretical foundation and practical solutions for real-world applications in privacy-sensitive scenarios by decoding buttock thermal patterns. The dataset and code supporting this study are publicly available at: <span><span>https://github.com/AJ-1995/Thermal-Memory-of-Chair</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103331"},"PeriodicalIF":3.4,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-28DOI: 10.1016/j.displa.2025.103334
Mingyang Yu , Haotian Lu , Donglin Wang , Ji Du , Desheng Kong , Xiaoxuan Xu , Jing Xu
Lupus Nephritis (LN), a severe complication of Systemic Lupus Erythematosus (SLE), critically affects renal function. To improve diagnostic accuracy, multi-threshold image segmentation (MTIS) techniques based on metaheuristic (MH) algorithms are widely adopted. However, traditional MH algorithms often suffer from premature convergence, limiting their global search capabilities. This study proposes a Quantum-Enhanced Hybrid Gold Rush Optimizer (QHGRO) that integrates quantum computing to enhance optimization performance. QHGRO is applied to an MTIS framework that utilizes a non-local means two-dimensional histogram to encode image information and employs Rényi entropy as the fitness function. The optimizer incorporates a Quantum Computing-Driven Adaptive Variation strategy, where quantum superposition enables parallel exploration of multiple states, and quantum mutation introduces controlled randomness to enhance global search and avoid local optima. To further improve performance, QHGRO includes a Stochastic Lévy Flight strategy during the Collaboration between Prospectors phase to enhance exploration and population diversity, and a Dynamic Fitness Distance Balance strategy during the Gold Mining phase to improve convergence accuracy. Experimental results on CEC2017 and CEC2022 benchmark functions demonstrate that QHGRO achieves competitive performance, often approaching global optima. In two engineering design problems—Speed Reducer Design and Three-Bar Truss Design—QHGRO outperforms classical algorithms (PSO, GWO, DE), newer algorithms (NRBO, CPO, RUN, BKA, SBOA, GRO), and advanced variants (MPSO, IAGWO). In LN pathological image segmentation tasks, the proposed method generates clear, high-quality segmented images, offering valuable support for clinical diagnosis.
{"title":"Quantum-enhanced gold rush Optimizer for multi-threshold segmentation of lupus nephritis pathological images","authors":"Mingyang Yu , Haotian Lu , Donglin Wang , Ji Du , Desheng Kong , Xiaoxuan Xu , Jing Xu","doi":"10.1016/j.displa.2025.103334","DOIUrl":"10.1016/j.displa.2025.103334","url":null,"abstract":"<div><div>Lupus Nephritis (LN), a severe complication of Systemic Lupus Erythematosus (SLE), critically affects renal function. To improve diagnostic accuracy, multi-threshold image segmentation (MTIS) techniques based on metaheuristic (MH) algorithms are widely adopted. However, traditional MH algorithms often suffer from premature convergence, limiting their global search capabilities. This study proposes a Quantum-Enhanced Hybrid Gold Rush Optimizer (QHGRO) that integrates quantum computing to enhance optimization performance. QHGRO is applied to an MTIS framework that utilizes a non-local means two-dimensional histogram to encode image information and employs Rényi entropy as the fitness function. The optimizer incorporates a Quantum Computing-Driven Adaptive Variation strategy, where quantum superposition enables parallel exploration of multiple states, and quantum mutation introduces controlled randomness to enhance global search and avoid local optima. To further improve performance, QHGRO includes a Stochastic Lévy Flight strategy during the Collaboration between Prospectors phase to enhance exploration and population diversity, and a Dynamic Fitness Distance Balance strategy during the Gold Mining phase to improve convergence accuracy. Experimental results on CEC2017 and CEC2022 benchmark functions demonstrate that QHGRO achieves competitive performance, often approaching global optima. In two engineering design problems—Speed Reducer Design and Three-Bar Truss Design—QHGRO outperforms classical algorithms (PSO, GWO, DE), newer algorithms (NRBO, CPO, RUN, BKA, SBOA, GRO), and advanced variants (MPSO, IAGWO). In LN pathological image segmentation tasks, the proposed method generates clear, high-quality segmented images, offering valuable support for clinical diagnosis.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103334"},"PeriodicalIF":3.4,"publicationDate":"2025-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145883334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-28DOI: 10.1016/j.displa.2025.103335
Bo Wang , Yuan Chen , Ayin Yan , Kepan Xu , Wenhao Bao , Yu Luo , Wenqing Xie , Xinshuo Zhang , Ying He
The proliferation of glass curtain wall LED media facade displays (G-LMDs) is transforming urban night environments but also introducing significant visual discomfort to observers and contributing to light pollution through glare, sky glow, and light intrusion. These impacts arise from the outward-facing luminous light sources mounted on building facades, which generate high luminance and strong luminance contrast at night. Existing building facade luminance standards, which were formulated for floodlighting, cannot evaluate and guide this new typology. This study proposes a novel “point-line-surface” luminance evaluation method that integrates the lighting characteristics of G-LMDs with human luminance perception properties. We quantified the human perceptual impact of G-LMDs by conducting luminance tests on their point and line sources from an observer’s perspective and converting the results into equivalent surface-source luminance. A key finding is that comfortable surface luminance is less influenced by array spacing and type, demonstrating high stability, which supports its use as a reliable metric for evaluating and controlling G-LMDs luminance. Based on this stability and its variability with ambient luminance, this study proposes G-LMDs luminance control values of 70 cd/m2, 65 cd/m2, and 50 cd/m2 for high, medium, and low ambient luminance, respectively. Furthermore, we invert the evaluation method into a practical “surface-line-point” design strategy to translate perceptual luminance targets into actionable lighting parameters, offering specific recommendations for different ambient luminance conditions. The proposed evaluation and control method and design strategy offer practical guidance for the design of urban G-LMDs and presents a viable strategy for mitigating urban light pollution and supporting landscape management.
{"title":"Luminance evaluation and control method for glass curtain wall LED media facade displays based on human visual perception","authors":"Bo Wang , Yuan Chen , Ayin Yan , Kepan Xu , Wenhao Bao , Yu Luo , Wenqing Xie , Xinshuo Zhang , Ying He","doi":"10.1016/j.displa.2025.103335","DOIUrl":"10.1016/j.displa.2025.103335","url":null,"abstract":"<div><div>The proliferation of glass curtain wall LED media facade displays (G-LMDs) is transforming urban night environments but also introducing significant visual discomfort to observers and contributing to light pollution through glare, sky glow, and light intrusion. These impacts arise from the outward-facing luminous light sources mounted on building facades, which generate high luminance and strong luminance contrast at night. Existing building facade luminance standards, which were formulated for floodlighting, cannot evaluate and guide this new typology. This study proposes a novel “point-line-surface” luminance evaluation method that integrates the lighting characteristics of G-LMDs with human luminance perception properties. We quantified the human perceptual impact of G-LMDs by conducting luminance tests on their point and line sources from an observer’s perspective and converting the results into equivalent surface-source luminance. A key finding is that comfortable surface luminance is less influenced by array spacing and type, demonstrating high stability, which supports its use as a reliable metric for evaluating and controlling G-LMDs luminance. Based on this stability and its variability with ambient luminance, this study proposes G-LMDs luminance control values of 70 cd/m<sup>2</sup>, 65 cd/m<sup>2</sup>, and 50 cd/m<sup>2</sup> for high, medium, and low ambient luminance, respectively. Furthermore, we invert the evaluation method into a practical “surface-line-point” design strategy to translate perceptual luminance targets into actionable lighting parameters, offering specific recommendations for different ambient luminance conditions. The proposed evaluation and control method and design strategy offer practical guidance for the design of urban G-LMDs and presents a viable strategy for mitigating urban light pollution and supporting landscape management.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103335"},"PeriodicalIF":3.4,"publicationDate":"2025-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145883335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-27DOI: 10.1016/j.displa.2025.103333
Jiajia Wang, Xiwu Shang, Peizhi Cheng, Guoping Li, Guozhong Wang
With the increasing prevalence of video coding standards like H.266/VVC, high-compression videos suffer from quality degradation including blocking artifacts and chroma blurring. Existing deep learning-based quality enhancement approaches primarily focus on spatial features, often overlooking global structural information in the frequency domain, which limits their ability to address complex compression distortions. To overcome these limitations, this study proposes a Multi-Scale Transformer and Frequency-Spatial Fusion Network (MSTF-Net), a CNN-Transformer hybrid architecture. MSTF-Net employs a collaborative optimization mechanism where Fourier-transformed frequency features guide spatial feature compensation. Furthermore, a dual-weighting mechanism is introduced to align the enhancement with human perception. This strategy enables the network to prioritize the enhancement of regions highly sensitive to human vision and suppresses redundant processing in homogeneous areas. Experimental results demonstrate that MSTF-Net achieves average improvements of 1.121 dB, 0.0112 and 0.36 in PSNR, SSIM and VMAF, which confirms its effectiveness for intra-frame compressed quality enhancement.
{"title":"MSTF-Net: A Multi-Scale Transformer and Frequency-Spatial Fusion Network for compressed video frame quality enhancement (ChinaMM)","authors":"Jiajia Wang, Xiwu Shang, Peizhi Cheng, Guoping Li, Guozhong Wang","doi":"10.1016/j.displa.2025.103333","DOIUrl":"10.1016/j.displa.2025.103333","url":null,"abstract":"<div><div>With the increasing prevalence of video coding standards like H.266/VVC, high-compression videos suffer from quality degradation including blocking artifacts and chroma blurring. Existing deep learning-based quality enhancement approaches primarily focus on spatial features, often overlooking global structural information in the frequency domain, which limits their ability to address complex compression distortions. To overcome these limitations, this study proposes a Multi-Scale Transformer and Frequency-Spatial Fusion Network (MSTF-Net), a CNN-Transformer hybrid architecture. MSTF-Net employs a collaborative optimization mechanism where Fourier-transformed frequency features guide spatial feature compensation. Furthermore, a dual-weighting mechanism is introduced to align the enhancement with human perception. This strategy enables the network to prioritize the enhancement of regions highly sensitive to human vision and suppresses redundant processing in homogeneous areas. Experimental results demonstrate that MSTF-Net achieves average improvements of 1.121 dB, 0.0112 and 0.36 in PSNR, SSIM and VMAF, which confirms its effectiveness for intra-frame compressed quality enhancement.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103333"},"PeriodicalIF":3.4,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145883336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-26DOI: 10.1016/j.displa.2025.103330
Shaocong Mo , Ming Cai , Lanfen Lin , Ruofeng Tong , Fang Wang , Qingqing Chen , Yinhao Li , Hongjie Hu , Yen-Wei Chen
Leveraging language semantics to strengthen vision information has become an advancing research in medical image segmentation. However, most existing studies restrict vision-language interactions in pair-wise correlations and only employ full-parameter fine-tuning under a single fusion scheme, leading to semantic inconsistency and potentiality ignorance of pre-trained medical vision-language models. In this paper, we introduce a parameter-efficient multimodal hypergraph adapter based model under a hybrid fusion paradigm of bidirectional early-fusion and late-fusion. Specifically, the multimodal hypergraph adapter is proposed to enhance interactions between several stages (blocks) of the vision and language encoders in medical CLIP models, which exploits high-order correlations by treating visual features as nodes and textual features as hyperedges, and propagates complementary information to refine features concurrently through hypergraph message passing. Additionally, we explore the hybrid utilization of word- and sentence-level textual features for fine-grained alignments in the task-specific decoder. Furthermore, two training strategies are employed for encoders from medical CLIP models: keeping the language encoder frozen only, or keeping both encoders frozen as parameter-efficient fine-tuning (PEFT) to fully exploit prior knowledge in the pre-trained model via adapters. Our experimental evaluations on the QaTa-COV19, CheXlocalize and MosMedData+ datasets demonstrate that our method significantly surpasses existing approaches, highlighting its superiority and effectiveness for text-driven medical image segmentation.
{"title":"MHG-Adapter: A parameter-efficient multimodal hypergraph adapter for text-driven medical image segmentation","authors":"Shaocong Mo , Ming Cai , Lanfen Lin , Ruofeng Tong , Fang Wang , Qingqing Chen , Yinhao Li , Hongjie Hu , Yen-Wei Chen","doi":"10.1016/j.displa.2025.103330","DOIUrl":"10.1016/j.displa.2025.103330","url":null,"abstract":"<div><div>Leveraging language semantics to strengthen vision information has become an advancing research in medical image segmentation. However, most existing studies restrict vision-language interactions in pair-wise correlations and only employ full-parameter fine-tuning under a single fusion scheme, leading to semantic inconsistency and potentiality ignorance of pre-trained medical vision-language models. In this paper, we introduce a parameter-efficient multimodal hypergraph adapter based model under a hybrid fusion paradigm of bidirectional early-fusion and late-fusion. Specifically, the multimodal hypergraph adapter is proposed to enhance interactions between several stages (blocks) of the vision and language encoders in medical CLIP models, which exploits high-order correlations by treating visual features as nodes and textual features as hyperedges, and propagates complementary information to refine features concurrently through hypergraph message passing. Additionally, we explore the hybrid utilization of word- and sentence-level textual features for fine-grained alignments in the task-specific decoder. Furthermore, two training strategies are employed for encoders from medical CLIP models: keeping the language encoder frozen only, or keeping both encoders frozen as parameter-efficient fine-tuning (PEFT) to fully exploit prior knowledge in the pre-trained model via adapters. Our experimental evaluations on the QaTa-COV19, CheXlocalize and MosMedData+ datasets demonstrate that our method significantly surpasses existing approaches, highlighting its superiority and effectiveness for text-driven medical image segmentation.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103330"},"PeriodicalIF":3.4,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145883474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-24DOI: 10.1016/j.displa.2025.103332
Long Qian , Yilin Chen , Yuxuan Hong , Lizhuang Ma , Xiao Lin
Single-image super-resolution (SISR) plays an important role in a wide range of computer vision applications. Although recent SISR methods have achieved good results, there is no effective solution to the problem of high-frequency feature loss after super-resolution due to attentional drift caused by deep attention architectures. To address these problems, we propose a novel Transformer-based network called RFASR, which pioneers the design of a global attention boosting module (PixBoost) and a Self-Graph Attention (SGA) upsampling module. The former enhances feature representation by combining meta-learning and multi-spatial channel aggregation, while the latter improves the modeling of high-frequency details through graph-based attention in the up-sampling phase. Experimental results on multiple datasets show that RFASR has a very high reconstruction level while maintaining optimal recovery efficiency.
{"title":"Perceptually robust super-resolution through global feature awareness","authors":"Long Qian , Yilin Chen , Yuxuan Hong , Lizhuang Ma , Xiao Lin","doi":"10.1016/j.displa.2025.103332","DOIUrl":"10.1016/j.displa.2025.103332","url":null,"abstract":"<div><div>Single-image super-resolution (SISR) plays an important role in a wide range of computer vision applications. Although recent SISR methods have achieved good results, there is no effective solution to the problem of high-frequency feature loss after super-resolution due to attentional drift caused by deep attention architectures. To address these problems, we propose a novel Transformer-based network called RFASR, which pioneers the design of a global attention boosting module (PixBoost) and a Self-Graph Attention (SGA) upsampling module. The former enhances feature representation by combining meta-learning and multi-spatial channel aggregation, while the latter improves the modeling of high-frequency details through graph-based attention in the up-sampling phase. Experimental results on multiple datasets show that RFASR has a very high reconstruction level while maintaining optimal recovery efficiency.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103332"},"PeriodicalIF":3.4,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145883475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the field of laser display technology, the spatial light modulator (SLM) functions as the core device for achieving display. Nevertheless, due to its inherent physical characteristics, one single spatial light modulator system has an upper limit on contrast ratio, making it arduous to meet the requirements of high dynamic range (HDR) display. Therefore, two spatial light modulators are frequently utilized to extend the dynamic range, and the level of pixel alignment between the two devices emerges as the decisive factor for system performance. In this research, a method based on double-slit interference was proposed to achieve pixel alignment between phase-only and amplitude-only spatial light modulators, and a high dynamic range laser display system based on the phase light modulator (PLM) and digital micromirror (DMD) was constructed. By calculating the relative position of the interference fringes to reveal the relative alignment state of the PLM and the DMD in the laser light path, including the translation and the inclination angles in three dimensions. Experiments results demonstrate that the method is simple to operate and can meet the pixel alignment accuracy of PLM and DMD within 21.6 μm.
{"title":"Pixel alignment method with double-slit interference for high-dynamic-range laser display system","authors":"Dabo Guo , Yunchuan Chen , Bin Guo , Jia Yu , Guang Yuan","doi":"10.1016/j.displa.2025.103329","DOIUrl":"10.1016/j.displa.2025.103329","url":null,"abstract":"<div><div>In the field of laser display technology, the spatial light modulator (SLM) functions as the core device for achieving display. Nevertheless, due to its inherent physical characteristics, one single spatial light modulator system has an upper limit on contrast ratio, making it arduous to meet the requirements of high dynamic range (HDR) display. Therefore, two spatial light modulators are frequently utilized to extend the dynamic range, and the level of pixel alignment between the two devices emerges as the decisive factor for system performance. In this research, a method based on double-slit interference was proposed to achieve pixel alignment between phase-only and amplitude-only spatial light modulators, and a high dynamic range laser display system based on the phase light modulator (PLM) and digital micromirror (DMD) was constructed. By calculating the relative position of the interference fringes to reveal the relative alignment state of the PLM and the DMD in the laser light path, including the translation and the inclination angles in three dimensions. Experiments results demonstrate that the method is simple to operate and can meet the pixel alignment accuracy of PLM and DMD within 21.6 μm.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103329"},"PeriodicalIF":3.4,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145839193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-20DOI: 10.1016/j.displa.2025.103327
Stephen J. Fieffer , Jonathan W. Kelly , Michael C. Dorneich , Stephen B. Gilbert
Visually induced motion sickness (VIMS) and motion sickness (MS) are frequent issues that occur when using screens or moving. While several individual characteristics influence one’s susceptibility to VIMS and MS, one of the more commonly cited factors is video game experience. However, gaming experience is typically considered as a binary variable and its impact has not been systematically characterized. This study employed a large sample (n = 1216) to examine the relationships between components of gaming experience and sickness susceptibility. The Gaming Activity and Motion-sickness Experience and Susceptibility Survey (GAMESS) was created to assess an individual’s status as current and previous gamer, age of first gameplay, and characteristics of typical game play: duration, graphical game type, and game genre. GAMESS also includes brief demographics of age, gender, race, and education, and the short VIMS susceptibility questionnaire (VIMSSQ) and MS susceptibility questionnaire (MSSQ). Current and previous gamers reported significantly lower sickness susceptibility than those who were never gamers, and gamers who played certain game genres (especially first-person shooter and battle royale) reported lower sickness susceptibility than those who played other genres. However, gender differences in sickness susceptibility suggests that gender is a critical factor in the relationship between gaming history and sickness susceptibility. These results highlight the importance of specific details of an individual’s gaming habits and history, going beyond the standard binary classification of gamers and non-gamers to include graphical styles and genres. This research may enable designers of simulated environments to better anticipate the sickness experienced by their users.
{"title":"The effect of video game history components on virtual and motion sickness susceptibility","authors":"Stephen J. Fieffer , Jonathan W. Kelly , Michael C. Dorneich , Stephen B. Gilbert","doi":"10.1016/j.displa.2025.103327","DOIUrl":"10.1016/j.displa.2025.103327","url":null,"abstract":"<div><div>Visually induced motion sickness (VIMS) and motion sickness (MS) are frequent issues that occur when using screens or moving. While several individual characteristics influence one’s susceptibility to VIMS and MS, one of the more commonly cited factors is video game experience. However, gaming experience is typically considered as a binary variable and its impact has not been systematically characterized. This study employed a large sample (<em>n</em> = 1216) to examine the relationships between components of gaming experience and sickness susceptibility. The Gaming Activity and Motion-sickness Experience and Susceptibility Survey (GAMESS) was created to assess an individual’s status as current and previous gamer, age of first gameplay, and characteristics of typical game play: duration, graphical game type, and game genre. GAMESS also includes brief demographics of age, gender, race, and education, and the short VIMS susceptibility questionnaire (VIMSSQ) and MS susceptibility questionnaire (MSSQ). Current and previous gamers reported significantly lower sickness susceptibility than those who were never gamers, and gamers who played certain game genres (especially first-person shooter and battle royale) reported lower sickness susceptibility than those who played other genres. However, gender differences in sickness susceptibility suggests that gender is a critical factor in the relationship between gaming history and sickness susceptibility. These results highlight the importance of specific details of an individual’s gaming habits and history, going beyond the standard binary classification of gamers and non-gamers to include graphical styles and genres. This research may enable designers of simulated environments to better anticipate the sickness experienced by their users.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103327"},"PeriodicalIF":3.4,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145839194","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-18DOI: 10.1016/j.displa.2025.103323
Luyang Xiao , Yixiao Liu , Xiao Liu , Hong Yang , Yuanyuan Wu , Chao Ren
Recovering missing details in low-resolution (LR) images with unknown degradations is the main challenge for real-world image super-resolution (Real-ISR) task. Nevertheless, recovering all types of these unknown degradations is usually too complex by using only one specific model. In the study, we find that the degradations of different real-world images have both commonalities and specificities. Therefore, we propose a brand-new Mixture-of-Degradation-Experts (MoDE) Transformer network for dealing with the commonalities and specificities in degraded images. To process the commonalities of LR images, we set MoDE blocks with identical structure in different depth of our network. To process the specificities of LR images, there are a number of experts in every MoDE block with different parameters learned by the network adaptively. These experts excel in dealing with different types of degradations, and our network assigns the most appropriate expert for different images with specific degradations guided by our proposed degradation representation feature extraction branch. Consequently, the collaboration between different experts in different depth of our network complete the Real-ISR task with complex and diverse degradation images. Our approach shows good performance compared to current state-of-the-arts (SOTA) methods by conducting extensive experiments.
{"title":"Degradation-Aware Mixture-of-Experts for Real-World Image Super-Resolution","authors":"Luyang Xiao , Yixiao Liu , Xiao Liu , Hong Yang , Yuanyuan Wu , Chao Ren","doi":"10.1016/j.displa.2025.103323","DOIUrl":"10.1016/j.displa.2025.103323","url":null,"abstract":"<div><div>Recovering missing details in low-resolution (LR) images with unknown degradations is the main challenge for real-world image super-resolution (Real-ISR) task. Nevertheless, recovering all types of these unknown degradations is usually too complex by using only one specific model. In the study, we find that the degradations of different real-world images have both commonalities and specificities. Therefore, we propose a brand-new Mixture-of-Degradation-Experts (MoDE) Transformer network for dealing with the commonalities and specificities in degraded images. To process the commonalities of LR images, we set MoDE blocks with identical structure in different depth of our network. To process the specificities of LR images, there are a number of experts in every MoDE block with different parameters learned by the network adaptively. These experts excel in dealing with different types of degradations, and our network assigns the most appropriate expert for different images with specific degradations guided by our proposed degradation representation feature extraction branch. Consequently, the collaboration between different experts in different depth of our network complete the Real-ISR task with complex and diverse degradation images. Our approach shows good performance compared to current state-of-the-arts (SOTA) methods by conducting extensive experiments.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"92 ","pages":"Article 103323"},"PeriodicalIF":3.4,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145839196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号