Low-light image enhancement(LLIE)remains challenging due to underexposure,color distortion,and amplified noise introduced during illumination correction.Existing deep learning–based methods typically apply uniform en...Low-light image enhancement(LLIE)remains challenging due to underexposure,color distortion,and amplified noise introduced during illumination correction.Existing deep learning–based methods typically apply uniform enhancement across the entire image,which overlooks scene semantics and often leads to texture degradation or unnatural color reproduction.To overcome these limitations,we propose a Semantic-Guided Visual Mamba Network(SGVMNet)that unifies semantic reasoning,state-space modeling,and mixture-of-experts routing for adaptive illumination correction.SGVMNet comprises three key components:(1)a semantic modulation module(SMM)that extracts scene-aware semantic priors from pretrained multimodal models—Large Language and Vision Assistant(LLaVA)and Contrastive Language–Image Pretraining(CLIP)—and injects them hierarchically into the feature stream;(2)aMixture-of-Experts State-Space Feature EnhancementModule(MoE-SSMFEM)that dynamically selects informative channels and activates specialized state-space experts for efficient global–local illumination modeling;and(3)a Text-Guided Mixture Mamba Block(TGMB)that fuses semantic priors and visual features through bidirectional state propagation.Experimental results demonstrate that on the low-light(LOL)dataset,SGVMNet outperforms other state-of-the-art methods in both quantitative and qualitative evaluations,and it also maintains low computational complexity with fast inference speed.On LOLv2-Syn,SGVMNet achieves 26.512 dB PSNR and 0.935 SSIM,outperforming RetinexFormer by 0.61 dB.On LOLv1,SGVMNet attains 26.50 dB PSNR and 0.863 SSIM.Furthermore,experiments on multiple unpaired real-world datasets further validate the superiority of SGVMNet,showing that the model not only exhibits strong cross-scene generalization ability but also effectively preserves semantic consistency and visual naturalness.展开更多
OBJECTIVE： To investigate the potential rules and knowledge of Traditional Chinese Medicine （TCM） and Western Medicine （WM） treatment on chronic urticaria （CU） based on data-mining methods. METHODS： Sixty pati...OBJECTIVE： To investigate the potential rules and knowledge of Traditional Chinese Medicine （TCM） and Western Medicine （WM） treatment on chronic urticaria （CU） based on data-mining methods. METHODS： Sixty patients with chronic urticaria, treated with TCM and WM, were selected. Gray correlation analyses were adopted to determine therapeutic efficacy. Association algorithms were utilized to ascertain the correlation between the disease course and treatment results. A genetic algorithm was applied to discover the optimization model in theTCM and WM treatment on CU. RESULTS： The total symptom scores after 4 weeks and 8 weeks of treatment in the TCM spleen-strengthening group correlated highly with the pretreatment total symptom score. The duration of treatment showed the greatest impact on the total symptom score. A quartic equation was established （y= - 1.6403x 10 - 6x4＋0.00025576x3＋0.0012819 x2 - 1.024x＋79.5879, and x=106.9518, y=83.0036） using the genetic algorithm. CONCLUSION： TCM treatment had a better effect in the later stage, whereas WM was better in the early stage. The duration of disease course had an impact on the effects of treatment. If the average total symptom score before treatment was 〈 83.0036, TCM or WM treatment could achieve better efficacy.展开更多
Sulfidation of zero-valent iron(ZVI)has attracted broad attention in recent years for improving the sequestration of contaminants from water.However,sulfidated ZVI(S-ZVI)is mostly synthesized in the aqueous phase,whic...Sulfidation of zero-valent iron(ZVI)has attracted broad attention in recent years for improving the sequestration of contaminants from water.However,sulfidated ZVI(S-ZVI)is mostly synthesized in the aqueous phase,which usually causes the formation of a thick iron oxide layer on the ZVI surface and hinders the efficient electron transfer to the contaminants.In this study,an alcohothermal strategy was employed for S-ZVI synthesis by the one-step reaction of iron powder with elemental sulfur.It is found that ferrous sulfide(FeS)with high purity and fine crystallization was formed on the ZVI surface,which is extremely favorable for electron transfer.Cr(Ⅵ)removal experiments confirm that the rate constant of SZVI synthesized by the alcohothermal method was 267.1-and 5.4-fold higher than those of un-sulfidated ZVI and aqueous-phase synthesized S-ZVI,respectively.Systematic characterizations proved that Cr(Ⅵ)was reduced and co-precipitated on S-ZVI in the form of a Fe(Ⅲ)/Cr(Ⅲ)/Cr(Ⅵ)composite,suggesting its environmental benignancy.展开更多
Wireless relay and network coding are two critical techniques to increase the reliability and throughput of wireless cooperative communication systems. In this paper, a complex field network coding （CFNC） scheme wit...Wireless relay and network coding are two critical techniques to increase the reliability and throughput of wireless cooperative communication systems. In this paper, a complex field network coding （CFNC） scheme with the K-th best relay selection （KBS） is proposed and investigated, wherein the K-th best relay is selected to forward the multiplexed signal to the destination. First, the upper bound of the symbol error probability （SEP）, the diversity order, and the coding gain are derived for the CFNC scheme with KBS. Then, the coding gain is utilized as the optimized cri- terion to determine the optimal power allocation. It is validated through analysis and simulation that the CFNC scheme with KBS can achieve full diversity only when K=I, while the diversity order decreases with increasing parameter K, and the optimal power allocation can significantly improve the performance of the CFNC scheme with KBS.展开更多
文摘Low-light image enhancement(LLIE)remains challenging due to underexposure,color distortion,and amplified noise introduced during illumination correction.Existing deep learning–based methods typically apply uniform enhancement across the entire image,which overlooks scene semantics and often leads to texture degradation or unnatural color reproduction.To overcome these limitations,we propose a Semantic-Guided Visual Mamba Network(SGVMNet)that unifies semantic reasoning,state-space modeling,and mixture-of-experts routing for adaptive illumination correction.SGVMNet comprises three key components:(1)a semantic modulation module(SMM)that extracts scene-aware semantic priors from pretrained multimodal models—Large Language and Vision Assistant(LLaVA)and Contrastive Language–Image Pretraining(CLIP)—and injects them hierarchically into the feature stream;(2)aMixture-of-Experts State-Space Feature EnhancementModule(MoE-SSMFEM)that dynamically selects informative channels and activates specialized state-space experts for efficient global–local illumination modeling;and(3)a Text-Guided Mixture Mamba Block(TGMB)that fuses semantic priors and visual features through bidirectional state propagation.Experimental results demonstrate that on the low-light(LOL)dataset,SGVMNet outperforms other state-of-the-art methods in both quantitative and qualitative evaluations,and it also maintains low computational complexity with fast inference speed.On LOLv2-Syn,SGVMNet achieves 26.512 dB PSNR and 0.935 SSIM,outperforming RetinexFormer by 0.61 dB.On LOLv1,SGVMNet attains 26.50 dB PSNR and 0.863 SSIM.Furthermore,experiments on multiple unpaired real-world datasets further validate the superiority of SGVMNet,showing that the model not only exhibits strong cross-scene generalization ability but also effectively preserves semantic consistency and visual naturalness.
文摘OBJECTIVE： To investigate the potential rules and knowledge of Traditional Chinese Medicine （TCM） and Western Medicine （WM） treatment on chronic urticaria （CU） based on data-mining methods. METHODS： Sixty patients with chronic urticaria, treated with TCM and WM, were selected. Gray correlation analyses were adopted to determine therapeutic efficacy. Association algorithms were utilized to ascertain the correlation between the disease course and treatment results. A genetic algorithm was applied to discover the optimization model in theTCM and WM treatment on CU. RESULTS： The total symptom scores after 4 weeks and 8 weeks of treatment in the TCM spleen-strengthening group correlated highly with the pretreatment total symptom score. The duration of treatment showed the greatest impact on the total symptom score. A quartic equation was established （y= - 1.6403x 10 - 6x4＋0.00025576x3＋0.0012819 x2 - 1.024x＋79.5879, and x=106.9518, y=83.0036） using the genetic algorithm. CONCLUSION： TCM treatment had a better effect in the later stage, whereas WM was better in the early stage. The duration of disease course had an impact on the effects of treatment. If the average total symptom score before treatment was 〈 83.0036, TCM or WM treatment could achieve better efficacy.
基金the National Key Research and Development Program of China(No.2019YFC1806203)for financial support。
文摘Sulfidation of zero-valent iron(ZVI)has attracted broad attention in recent years for improving the sequestration of contaminants from water.However,sulfidated ZVI(S-ZVI)is mostly synthesized in the aqueous phase,which usually causes the formation of a thick iron oxide layer on the ZVI surface and hinders the efficient electron transfer to the contaminants.In this study,an alcohothermal strategy was employed for S-ZVI synthesis by the one-step reaction of iron powder with elemental sulfur.It is found that ferrous sulfide(FeS)with high purity and fine crystallization was formed on the ZVI surface,which is extremely favorable for electron transfer.Cr(Ⅵ)removal experiments confirm that the rate constant of SZVI synthesized by the alcohothermal method was 267.1-and 5.4-fold higher than those of un-sulfidated ZVI and aqueous-phase synthesized S-ZVI,respectively.Systematic characterizations proved that Cr(Ⅵ)was reduced and co-precipitated on S-ZVI in the form of a Fe(Ⅲ)/Cr(Ⅲ)/Cr(Ⅵ)composite,suggesting its environmental benignancy.
基金supported by the Major State Basic Research Development Program of China(973 Program No.2012CB316100)the National Natural Science Foundation of China(Nos.61032002/61271246)the 111 Project(No.111-2-14)
文摘Wireless relay and network coding are two critical techniques to increase the reliability and throughput of wireless cooperative communication systems. In this paper, a complex field network coding （CFNC） scheme with the K-th best relay selection （KBS） is proposed and investigated, wherein the K-th best relay is selected to forward the multiplexed signal to the destination. First, the upper bound of the symbol error probability （SEP）, the diversity order, and the coding gain are derived for the CFNC scheme with KBS. Then, the coding gain is utilized as the optimized cri- terion to determine the optimal power allocation. It is validated through analysis and simulation that the CFNC scheme with KBS can achieve full diversity only when K=I, while the diversity order decreases with increasing parameter K, and the optimal power allocation can significantly improve the performance of the CFNC scheme with KBS.
