Pub Date : 2026-02-05DOI: 10.1016/j.displa.2026.103373
Yunjiao Ma, Zhenzhen He, Jun Xiang, Ning Zhang, Ruru Pan
Sketch colorization, using black-and-white lines for structure, faces inefficiency in manual methods, driving the need for intelligent solutions. Existing techniques, such as natural language, label-guided, and color-hint methods, suffer from spatial imprecision, limited color diversity, or poor modeling of long-range dependency and control of color change. The proposed Structural Hint-Guided Colorization Network generates color hints via superpixel decomposition with line density analysis to adapt to structural complexity, integrates a Transformer branch into the Residual Network for global dependencies, and uses a hybrid loss for controlled color transitions. Experiments show balanced precision and flexibility in complex sketches with a Peak Signal-to-Noise Ratio (PSNR) of 23.221 and a Structural Similarity Index Measure (SSIM) of 0.853. Compared with the baseline, an improvement of 5.63% PSNR and 3.90% SSIM validates the effectiveness of the proposed method.
{"title":"Structural hint-guided colorization network for sketch colorization","authors":"Yunjiao Ma, Zhenzhen He, Jun Xiang, Ning Zhang, Ruru Pan","doi":"10.1016/j.displa.2026.103373","DOIUrl":"10.1016/j.displa.2026.103373","url":null,"abstract":"<div><div>Sketch colorization, using black-and-white lines for structure, faces inefficiency in manual methods, driving the need for intelligent solutions. Existing techniques, such as natural language, label-guided, and color-hint methods, suffer from spatial imprecision, limited color diversity, or poor modeling of long-range dependency and control of color change. The proposed Structural Hint-Guided Colorization Network generates color hints via superpixel decomposition with line density analysis to adapt to structural complexity, integrates a Transformer branch into the Residual Network for global dependencies, and uses a hybrid loss for controlled color transitions. Experiments show balanced precision and flexibility in complex sketches with a Peak Signal-to-Noise Ratio (PSNR) of 23.221 and a Structural Similarity Index Measure (SSIM) of 0.853. Compared with the baseline, an improvement of 5.63% PSNR and 3.90% SSIM validates the effectiveness of the proposed method.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103373"},"PeriodicalIF":3.4,"publicationDate":"2026-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191718","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 : 2026-02-04DOI: 10.1016/j.displa.2026.103372
Chen Yang , Jixiang Nie , Hui Chen , Weina Wang , Wanquan Liu
Point cloud registration typically relies on point-pair feature extraction. However, point cloud features are low-dimensional, and point-wise processing lacks topological structure and leads to high computational complexity. Address to these challenges, a multi-view 3D point cloud registration method based on generated multi-scale information granules is proposed to build the completed 3D reconstruction. Specifically, during the granule generation process, Fast Persistent Feature Histograms (FPFH) are integrated into Fuzzy C-means clustering to ensure the preservation of geometric features while reducing computational cost. Furthermore, to ensure feature completeness across regions with varying densities, a surface complexity threshold is employed to merge fine-grained granules and eliminate relatively flat surfaces. This approach avoids over-segmentation and redundancy, thereby improving the efficiency of point cloud processing. Finally, to tackle the uneven distribution of overlapping areas and noise-induced mismatches, a hierarchical GMM-based 3D registration framework based on multi-scale information granules is constructed. Point cloud granules are dynamically updated in real time to ensure registration between granules with complete geometric features, thus improving registration accuracy. Experiments conducted on benchmark datasets and real-world collected data demonstrate that the proposed method outperforms existing methods in multi-view registration, offering improved accuracy and efficiency.
{"title":"Multi-view 3D point cloud registration method based on generated multi-scale information granules","authors":"Chen Yang , Jixiang Nie , Hui Chen , Weina Wang , Wanquan Liu","doi":"10.1016/j.displa.2026.103372","DOIUrl":"10.1016/j.displa.2026.103372","url":null,"abstract":"<div><div>Point cloud registration typically relies on point-pair feature extraction. However, point cloud features are low-dimensional, and point-wise processing lacks topological structure and leads to high computational complexity. Address to these challenges, a multi-view 3D point cloud registration method based on generated multi-scale information granules is proposed to build the completed 3D reconstruction. Specifically, during the granule generation process, Fast Persistent Feature Histograms (FPFH) are integrated into Fuzzy C-means clustering to ensure the preservation of geometric features while reducing computational cost. Furthermore, to ensure feature completeness across regions with varying densities, a surface complexity threshold is employed to merge fine-grained granules and eliminate relatively flat surfaces. This approach avoids over-segmentation and redundancy, thereby improving the efficiency of point cloud processing. Finally, to tackle the uneven distribution of overlapping areas and noise-induced mismatches, a hierarchical GMM-based 3D registration framework based on multi-scale information granules is constructed. Point cloud granules are dynamically updated in real time to ensure registration between granules with complete geometric features, thus improving registration accuracy. Experiments conducted on benchmark datasets and real-world collected data demonstrate that the proposed method outperforms existing methods in multi-view registration, offering improved accuracy and efficiency.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103372"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191253","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}
Generative Adversarial Networks (GAN) have significantly improved data security in image steganography. However, existing GAN-based approaches often fail to consider the impact of transmission noise and rely on separately trained encoder–decoder architectures, which hinder the accurate recovery of hidden image data. To address these limitations, we propose a Residual and Multi-Attention Enhanced GAN (RME-GAN) for image steganography, which integrates residual networks, attention mechanisms, and multi-objective optimization to effectively enhance the recovery quality of secret images. In the generator, a residual preprocessing network combined with a global attention mechanism is employed to efficiently extract transmission noise features. In the extractor, a gated attention module is introduced to align the encoder and decoder features, thereby improving decoding accuracy. Moreover, a multi-objective loss function is formulated to jointly optimize both encoder and decoder through end-to-end training, enhancing the consistency between them. Experimental results on widely used datasets, including LFW, ImageNet, and Pascal, demonstrate that the proposed RME-GAN achieves superior robustness against noise and significantly improves Peak Signal-to-Noise Ratio (PSNR) and Structure Similarity Index Measure (SSIM) performance compared to existing methods.
{"title":"Robust image steganography based on residual and multi-attention enhanced Generative Adversarial Networks","authors":"Yuling Luo, Zhaohui Chen, Baoshan Lu, Yiting Huang, Qiang Fu, Sheng Qin, Junxiu Liu","doi":"10.1016/j.displa.2026.103384","DOIUrl":"10.1016/j.displa.2026.103384","url":null,"abstract":"<div><div>Generative Adversarial Networks (GAN) have significantly improved data security in image steganography. However, existing GAN-based approaches often fail to consider the impact of transmission noise and rely on separately trained encoder–decoder architectures, which hinder the accurate recovery of hidden image data. To address these limitations, we propose a Residual and Multi-Attention Enhanced GAN (RME-GAN) for image steganography, which integrates residual networks, attention mechanisms, and multi-objective optimization to effectively enhance the recovery quality of secret images. In the generator, a residual preprocessing network combined with a global attention mechanism is employed to efficiently extract transmission noise features. In the extractor, a gated attention module is introduced to align the encoder and decoder features, thereby improving decoding accuracy. Moreover, a multi-objective loss function is formulated to jointly optimize both encoder and decoder through end-to-end training, enhancing the consistency between them. Experimental results on widely used datasets, including LFW, ImageNet, and Pascal, demonstrate that the proposed RME-GAN achieves superior robustness against noise and significantly improves Peak Signal-to-Noise Ratio (PSNR) and Structure Similarity Index Measure (SSIM) performance compared to existing methods.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103384"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191720","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 : 2026-02-04DOI: 10.1016/j.displa.2026.103379
Changrui Zhu , Ernst Kruijff , Harvey Stedman , Vijay M. Pawar , Simon Julier
Change detection is a cognitively challenging process that involves three stages: spotting (becoming aware of a change); localising (establishing the specific location of the change); and identifying (recognising the nature of the change). Each of these stages has the potential to be influenced by both the way the data is presented (e.g., display type) and the fidelity of that data. To explore these issues, we conducted two studies, both of which looked at the effects of display type (immersive virtual reality (VR) or desktop monitor (DM)), and the semantic availability of the scene (low or high realism).
Study 1 () explored the VR–DM differences in a broad scope, which examined six change types spanning both spatial and non-spatial changes—disappear, appear, translation, rotation, replacement, and colour. However, there were no significant differences between VR and DM in spotting, localising, and identifying at either level of (semantic) realism. Study 2 () followed this up by exploring only two types of spatial change (translation and rotation) at a much finer degree of granularity while retaining the same experimental paradigm with necessary refinement. Study 2 showed a significant VR advantage over DM, with different patterns across realism conditions: In low-realism scenes, VR significantly outperformed DM on localisation and change-type identification overall, with the largest VR–DM contrasts observed for the smallest translations. In high-realism scenes, the only significant effect was a display-by-magnitude interaction for change-type identification at the smallest translations. Taken both studies together, VR benefits are most likely for subtle spatial changes, particularly small translations, when the semantic availability is limited. Questionnaire ratings also suggested that reliance on visual features varies with semantic availability. Semantic cues were rated significantly higher than other features in high realism scenes only. Finally, there is no significant difference between VR and DM in terms of workload, motion sickness and self-confidence, suggesting that the perceptual advantages of VR come with no additional physical or cognitive costs for change detection.
{"title":"Does display type matter for change detection? comparing immersive and non-immersive displays under low and high semantic availability","authors":"Changrui Zhu , Ernst Kruijff , Harvey Stedman , Vijay M. Pawar , Simon Julier","doi":"10.1016/j.displa.2026.103379","DOIUrl":"10.1016/j.displa.2026.103379","url":null,"abstract":"<div><div>Change detection is a cognitively challenging process that involves three stages: spotting (becoming aware of a change); localising (establishing the specific location of the change); and identifying (recognising the nature of the change). Each of these stages has the potential to be influenced by both the way the data is presented (e.g., display type) and the fidelity of that data. To explore these issues, we conducted two studies, both of which looked at the effects of display type (immersive virtual reality (VR) or desktop monitor (DM)), and the semantic availability of the scene (low or high realism).</div><div>Study 1 (<span><math><mrow><mi>N</mi><mo>=</mo><mn>38</mn></mrow></math></span>) explored the VR–DM differences in a broad scope, which examined six change types spanning both spatial and non-spatial changes—<em>disappear</em>, <em>appear</em>, <em>translation</em>, <em>rotation</em>, <em>replacement</em>, and <em>colour</em>. However, there were no significant differences between VR and DM in spotting, localising, and identifying at either level of (semantic) realism. Study 2 (<span><math><mrow><mi>N</mi><mo>=</mo><mn>20</mn></mrow></math></span>) followed this up by exploring only two types of spatial change (<em>translation</em> and <em>rotation</em>) at a much finer degree of granularity while retaining the same experimental paradigm with necessary refinement. Study 2 showed a significant VR advantage over DM, with different patterns across realism conditions: In low-realism scenes, VR significantly outperformed DM on localisation and change-type identification overall, with the largest VR–DM contrasts observed for the smallest <em>translations</em>. In high-realism scenes, the only significant effect was a display-by-magnitude interaction for change-type identification at the smallest <em>translations</em>. Taken both studies together, VR benefits are most likely for subtle spatial changes, particularly small <em>translations</em>, when the semantic availability is limited. Questionnaire ratings also suggested that reliance on visual features varies with semantic availability. Semantic cues were rated significantly higher than other features in high realism scenes only. Finally, there is no significant difference between VR and DM in terms of workload, motion sickness and self-confidence, suggesting that the perceptual advantages of VR come with no additional physical or cognitive costs for change detection.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103379"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191722","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 : 2026-02-04DOI: 10.1016/j.displa.2026.103376
Yongqing Cai, Cheng Han, Wei Quan, Yuechen Zhang
With the annual increase in Virtual Reality (VR) products and content, an increasing number of users are engaging with VR videos. However, many users experience discomfort such as headaches and dizziness during VR experiences, a phenomenon known as VR sickness. To enhance user comfort during VR experiences, this study proposes a VR sickness assessment model based on visual attention mechanisms, enabling automatic classification of VR content so users can select experiences suitable for their needs. The proposed model comprises an attention stream subnetwork, inspired by user attention mechanisms, and a motion stream subnetwork, jointly forming a dual-stream evaluation system. Leveraging a transformer architecture, the model establishes self-attention mechanisms over temporal and spatial sequences to capture their interdependent features. A multi-level fusion strategy is employed to extract low-level, high-level, and global features, while attention mechanisms adaptively integrate these multi-level features, achieving precise VR sickness assessment results. Experiments conducted on publicly available datasets demonstrate the effectiveness of the visual attention mechanism in improving model assessment accuracy. The model achieved 88.18% and 92.22% accuracy on two public datasets, respectively, representing a significant performance improvement compared to existing studies.
{"title":"A visual Attention-Based model for VR sickness assessment","authors":"Yongqing Cai, Cheng Han, Wei Quan, Yuechen Zhang","doi":"10.1016/j.displa.2026.103376","DOIUrl":"10.1016/j.displa.2026.103376","url":null,"abstract":"<div><div>With the annual increase in Virtual Reality (VR) products and content, an increasing number of users are engaging with VR videos. However, many users experience discomfort such as headaches and dizziness during VR experiences, a phenomenon known as VR sickness. To enhance user comfort during VR experiences, this study proposes a VR sickness assessment model based on visual attention mechanisms, enabling automatic classification of VR content so users can select experiences suitable for their needs. The proposed model comprises an attention stream subnetwork, inspired by user attention mechanisms, and a motion stream subnetwork, jointly forming a dual-stream evaluation system. Leveraging a transformer architecture, the model establishes self-attention mechanisms over temporal and spatial sequences to capture their interdependent features. A multi-level fusion strategy is employed to extract low-level, high-level, and global features, while attention mechanisms adaptively integrate these multi-level features, achieving precise VR sickness assessment results. Experiments conducted on publicly available datasets demonstrate the effectiveness of the visual attention mechanism in improving model assessment accuracy. The model achieved 88.18% and 92.22% accuracy on two public datasets, respectively, representing a significant performance improvement compared to existing studies.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103376"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191719","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 : 2026-02-04DOI: 10.1016/j.displa.2026.103381
Mengxue Yan , Zirui Wang , Zhenfeng Li , Peng Wang , Pang Wu , Xianxiang Chen , Lidong Du , Li Li , Hongbo Chang , Zhen Fang
Accurate segmentation of the arterial lumen in ultrasound images is crucial for clinical diagnosis and hemodynamic assessment, but is challenged by inherent image properties such as low contrast, artifacts, and surrounding tissues with similar morphology. These factors conjointly lead to significant localization ambiguity, which severely hampers the performance of segmentation models. To address this issue, we propose a novel localization knowledge-driven Segmentation (LKDS) framework, which guides accurate segmentation through explicit localization. The proposed framework first acquires robust localization knowledge through a Localization Prior Learning (LPL) process on a coarsely-annotated dataset, which is then efficiently transferred and adapted to target datasets via a few-shot pseudo-labeling strategy. Operationally, the LKDS framework generates a dynamic Localization Map (LM) for each image to explicitly guide a subsequent network in performing accurate segmentation. Extensive experiments on two distinct arterial ultrasound datasets show that our LKDS framework not only accelerates training convergence but also significantly outperforms state-of-the-art implicit segmentation methods. Our work demonstrates that explicitly incorporating localization knowledge is an effective strategy for significantly enhancing the performance of arterial segmentation.
{"title":"Localization knowledge-driven segmentation of arteries in ultrasound images","authors":"Mengxue Yan , Zirui Wang , Zhenfeng Li , Peng Wang , Pang Wu , Xianxiang Chen , Lidong Du , Li Li , Hongbo Chang , Zhen Fang","doi":"10.1016/j.displa.2026.103381","DOIUrl":"10.1016/j.displa.2026.103381","url":null,"abstract":"<div><div>Accurate segmentation of the arterial lumen in ultrasound images is crucial for clinical diagnosis and hemodynamic assessment, but is challenged by inherent image properties such as low contrast, artifacts, and surrounding tissues with similar morphology. These factors conjointly lead to significant localization ambiguity, which severely hampers the performance of segmentation models. To address this issue, we propose a novel localization knowledge-driven Segmentation (LKDS) framework, which guides accurate segmentation through explicit localization. The proposed framework first acquires robust localization knowledge through a Localization Prior Learning (LPL) process on a coarsely-annotated dataset, which is then efficiently transferred and adapted to target datasets via a few-shot pseudo-labeling strategy. Operationally, the LKDS framework generates a dynamic Localization Map (LM) for each image to explicitly guide a subsequent network in performing accurate segmentation. Extensive experiments on two distinct arterial ultrasound datasets show that our LKDS framework not only accelerates training convergence but also significantly outperforms state-of-the-art implicit segmentation methods. Our work demonstrates that explicitly incorporating localization knowledge is an effective strategy for significantly enhancing the performance of arterial segmentation.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103381"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191723","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 : 2026-02-04DOI: 10.1016/j.displa.2026.103371
Meng Liu , Lei Fan , Jiechun Lin , Zhenhao Song , Qi Li , Yingxiang Han , Dajiang Wang , Xiaofei Wang
Virtual reality (VR) games are increasingly applied in various fields, yet prolonged immersion can induce VR sickness. In unidirectional VR game modes, targets mainly appear in front of the player, requiring minimal head movement, whereas multidirectional modes present targets from multiple directions, demanding greater head and body rotation for a full 360° experience. Although VR-induced discomfort has been studied, the specific effects of different modes on binocular visual function and VR sickness remain insufficiently understood. This study examined how two different viewing modes of the VR game Beat Saber—unidirectional and multidirectional—affect binocular visual function and subjective symptoms. Thirty-three participants played each mode for 30 min using a head-mounted display. The Simulator Sickness Questionnaire (SSQ), Visual Fatigue Scale, and phoropter-based binocular visual function tests were conducted before and after gameplay. Significant increases were observed in total and subscale SSQ scores and visual fatigue scores after both modes, indicating that VR gaming induces adverse symptoms. The accommodative convergence to accommodation ratio (AC/A) decreased significantly after the multidirectional mode, suggesting greater effects on binocular accommodation and vergence in multidirectional mode. Near exophoria was negatively correlated with visual fatigue after both the unidirectional and multidirectional modes; accommodative response (AR) correlated positively with visual fatigue after the unidirectional mode; and negative relative accommodation (NRA) correlated negatively with visual fatigue after the multidirectional mode. These findings provide insights into how VR gameplay mode influences VR sickness, visual fatigue, and binocular visual function, supporting the development of VR design standards and user experience optimization.
{"title":"Effects of different viewing modes in virtual reality games on visual function parameters and subjective symptoms: A cross-sectional study","authors":"Meng Liu , Lei Fan , Jiechun Lin , Zhenhao Song , Qi Li , Yingxiang Han , Dajiang Wang , Xiaofei Wang","doi":"10.1016/j.displa.2026.103371","DOIUrl":"10.1016/j.displa.2026.103371","url":null,"abstract":"<div><div>Virtual reality (VR) games are increasingly applied in various fields, yet prolonged immersion can induce VR sickness. In unidirectional VR game modes, targets mainly appear in front of the player, requiring minimal head movement, whereas multidirectional modes present targets from multiple directions, demanding greater head and body rotation for a full 360° experience. Although VR-induced discomfort has been studied, the specific effects of different modes on binocular visual function and VR sickness remain insufficiently understood. This study examined how two different viewing modes of the VR game Beat Saber—unidirectional and multidirectional—affect binocular visual function and subjective symptoms. Thirty-three participants played each mode for 30 min using a head-mounted display. The Simulator Sickness Questionnaire (SSQ), Visual Fatigue Scale, and phoropter-based binocular visual function tests were conducted before and after gameplay. Significant increases were observed in total and subscale SSQ scores and visual fatigue scores after both modes, indicating that VR gaming induces adverse symptoms. The accommodative convergence to accommodation ratio (AC/A) decreased significantly after the multidirectional mode, suggesting greater effects on binocular accommodation and vergence in multidirectional mode. Near exophoria was negatively correlated with visual fatigue after both the unidirectional and multidirectional modes; accommodative response (AR) correlated positively with visual fatigue after the unidirectional mode; and negative relative accommodation (NRA) correlated negatively with visual fatigue after the multidirectional mode. These findings provide insights into how VR gameplay mode influences VR sickness, visual fatigue, and binocular visual function, supporting the development of VR design standards and user experience optimization.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103371"},"PeriodicalIF":3.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191196","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 : 2026-02-03DOI: 10.1016/j.displa.2026.103370
Shuaishuai Wang , Zhihua Wang , Peiquan Zheng , Yijian Zhou , Chengwu Wu , Taifu Lang , Xin Lin , Xin Wu , Caihong Yan , Qun Yan , Kaixin Zhang , Jie Sun
To fabricate high-quality bump arrays on Micro-LEDs, thereby enhance the yield and reliability of high-density Micro-LED devices. We employed plasma treatment on Micro-LED samples and added a specific concentration of surfactant to the electroless plating solution, which enhance Micro-LEDs surface and electroless plating solution wettability, creating a conducive environment for fabrication of bumps. In comparison to traditional nickel bump fabrication on high-density substrates, we successfully fabricated fast growth rate, high uniformity, excellent shear strength and low surface roughness of nickel bumps on Micro-LEDs via the synergistic effect of plasma treatment and wettability electroless plating, the bump growth rate, array uniformity and shear strength improved by 56.8%, 86% and 377%, respectively, while surface roughness decreased by 94.4%. This work provides a critical pathway for fabricating high-quality nickel bumps and enhancing the yield and reliability of highly integrated Micro-LED devices.
{"title":"Electroless plating of high-quality Ni microbumps for high-density micro-LEDs realized via surface plasma treatment and solution wettability enhancement","authors":"Shuaishuai Wang , Zhihua Wang , Peiquan Zheng , Yijian Zhou , Chengwu Wu , Taifu Lang , Xin Lin , Xin Wu , Caihong Yan , Qun Yan , Kaixin Zhang , Jie Sun","doi":"10.1016/j.displa.2026.103370","DOIUrl":"10.1016/j.displa.2026.103370","url":null,"abstract":"<div><div>To fabricate high-quality bump arrays on Micro-LEDs, thereby enhance the yield and reliability of high-density Micro-LED devices. We employed plasma treatment on Micro-LED samples and added a specific concentration of surfactant to the electroless plating solution, which enhance Micro-LEDs surface and electroless plating solution wettability, creating a conducive environment for fabrication of bumps. In comparison to traditional nickel bump fabrication on high-density substrates, we successfully fabricated fast growth rate, high uniformity, excellent shear strength and low surface roughness of nickel bumps on Micro-LEDs via the synergistic effect of plasma treatment and wettability electroless plating, the bump growth rate, array uniformity and shear strength improved by 56.8%, 86% and 377%, respectively, while surface roughness decreased by 94.4%. This work provides a critical pathway for fabricating high-quality nickel bumps and enhancing the yield and reliability of highly integrated Micro-LED devices.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103370"},"PeriodicalIF":3.4,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191705","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 : 2026-02-02DOI: 10.1016/j.displa.2026.103374
Yani Guo , Zhenhong Jia , Gang Zhou , Xiaohui Huang , Yue Li , Mingyan Li , Guohong Chen , Junjie Li
Numerous obstacles are faced in change detection tasks for large-field-of-view video images (e.g., those acquired by Eagle Eye devices) in low-light environments, mainly due to the difficulty in differentiating genuine changes from illumination-induced pseudo-changes, vulnerability to intricate noise interference, and constrained robustness in multi-scale change detection. To address these issues, a deep learning framework for large-field-of-view change detection in low-light environments is proposed in this paper, consisting of three core modules: Cross-scale Attention Feature Fusion, Difference Enhancement and Optimization, and Pseudo-Change Suppression and Multi-scale Fusion. Initially, the Cross-scale Attention Feature Fusion (CAF) module employs a cross-scale attention mechanism to fuse multi-scale features, capturing change information at various scales. Structural differences are then enhanced by the Difference Enhancement and Optimization (DEO) module through frequency-domain decomposition and boundary-aware strategies, mitigating the impact of illumination variations. Subsequently, illumination-induced pseudo-changes are suppressed by the Pseudo-Change Suppression and Multi-scale Fusion (PSF) module with Pseudo-Change Filtering Attention, and multi-scale feature fusion is performed to generate accurate change maps. Additionally, an end-to-end optimization strategy is introduced, incorporating contrastive learning and self-supervised pseudo-label generation, to further enhance the model’s robustness and generalization across various low-light scenarios. Experimental results demonstrate that, compared with other methods, The method described in this paper improved the F1 score by 3.65% and accuracy by 1.84%, verifying its ability to accurately distinguish between real and false changes in low-light environments.
{"title":"Change detection of large-field-of-view video images in low-light environments with cross-scale feature fusion and pseudo-change mitigation","authors":"Yani Guo , Zhenhong Jia , Gang Zhou , Xiaohui Huang , Yue Li , Mingyan Li , Guohong Chen , Junjie Li","doi":"10.1016/j.displa.2026.103374","DOIUrl":"10.1016/j.displa.2026.103374","url":null,"abstract":"<div><div>Numerous obstacles are faced in change detection tasks for large-field-of-view video images (e.g., those acquired by Eagle Eye devices) in low-light environments, mainly due to the difficulty in differentiating genuine changes from illumination-induced pseudo-changes, vulnerability to intricate noise interference, and constrained robustness in multi-scale change detection. To address these issues, a deep learning framework for large-field-of-view change detection in low-light environments is proposed in this paper, consisting of three core modules: Cross-scale Attention Feature Fusion, Difference Enhancement and Optimization, and Pseudo-Change Suppression and Multi-scale Fusion. Initially, the Cross-scale Attention Feature Fusion (CAF) module employs a cross-scale attention mechanism to fuse multi-scale features, capturing change information at various scales. Structural differences are then enhanced by the Difference Enhancement and Optimization (DEO) module through frequency-domain decomposition and boundary-aware strategies, mitigating the impact of illumination variations. Subsequently, illumination-induced pseudo-changes are suppressed by the Pseudo-Change Suppression and Multi-scale Fusion (PSF) module with Pseudo-Change Filtering Attention, and multi-scale feature fusion is performed to generate accurate change maps. Additionally, an end-to-end optimization strategy is introduced, incorporating contrastive learning and self-supervised pseudo-label generation, to further enhance the model’s robustness and generalization across various low-light scenarios. Experimental results demonstrate that, compared with other methods, The method described in this paper improved the F1 score by 3.65% and accuracy by 1.84%, verifying its ability to accurately distinguish between real and false changes in low-light environments.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103374"},"PeriodicalIF":3.4,"publicationDate":"2026-02-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191088","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 : 2026-02-02DOI: 10.1016/j.displa.2026.103377
Aizhong Zhou , Fengbo Wang , Jiong Guo , Yutao Liu
The Mixture of Experts (MoE) is a neural network architecture which is widely used in fields such as natural language processing (such as large language models, multilingual translation), computer vision (such as medical image analysis, multi-modal learning), and recommendation systems. A core problem of the MoE is how to select an expert assigned to a specific task among all experts. This problem can be transformed into an election problem where each expert is a candidate and the winner of election (a candidate or some candidates) is the expert who is assigned to the task by considering the votes. We study a variant of committee elections from the perspective of computational complexity. Given a set of candidates, each possessing a set of attributes and a profit value, and a set of constraints specified as propositional logical expressions on the attributes, the task is to select a committee of candidates that satisfies all constraints and whose total profit meets a given threshold. Regarding the classical complexity, we design two polynomial time algorithms for two special conditions and provide some NP-hardness results. Moreover, we examine the parameterized complexity and get some FPT, W[1]-hard and para-NP-hard results.
{"title":"Committee Elections with Candidate Attribute Constraints","authors":"Aizhong Zhou , Fengbo Wang , Jiong Guo , Yutao Liu","doi":"10.1016/j.displa.2026.103377","DOIUrl":"10.1016/j.displa.2026.103377","url":null,"abstract":"<div><div>The Mixture of Experts (MoE) is a neural network architecture which is widely used in fields such as natural language processing (such as large language models, multilingual translation), computer vision (such as medical image analysis, multi-modal learning), and recommendation systems. A core problem of the MoE is how to select an expert assigned to a specific task among all experts. This problem can be transformed into an election problem where each expert is a candidate and the winner of election (a candidate or some candidates) is the expert who is assigned to the task by considering the votes. We study a variant of committee elections from the perspective of computational complexity. Given a set of candidates, each possessing a set of attributes and a profit value, and a set of constraints specified as propositional logical expressions on the attributes, the task is to select a committee of <span><math><mi>k</mi></math></span> candidates that satisfies all constraints and whose total profit meets a given threshold. Regarding the classical complexity, we design two polynomial time algorithms for two special conditions and provide some NP-hardness results. Moreover, we examine the parameterized complexity and get some FPT, W[1]-hard and para-NP-hard results.</div></div>","PeriodicalId":50570,"journal":{"name":"Displays","volume":"93 ","pages":"Article 103377"},"PeriodicalIF":3.4,"publicationDate":"2026-02-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191193","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