Obtaining accurate bathymetric maps is very valuable for marine environment monitoring,port planning,and so on.Accurately estimating water depth in turbid coastal waters using satellite remote sensing encounters chall...Obtaining accurate bathymetric maps is very valuable for marine environment monitoring,port planning,and so on.Accurately estimating water depth in turbid coastal waters using satellite remote sensing encounters challenges originating from low water transparency,but it is limited by the quantity,quality,and water quality of samples.This study introduces a fast feature cascade learning model(FFCLM)to enhance the accuracy of bathymetric inversion from multispectral satellite images,particularly when limited field samples are available.FFCLM leverages spectral bands and in situ data to derive effective inversion weights through feature concatenation and cascade fitting.Field experiments conducted at Nanshan Port and Rushikonda Beach gathered water depth,satellite,and in situ data.Comparative analysis with conventional machine learning algorithms,including support vector machine,random forest,and gradient boosting trees,indicates that FFCLM achieves lower errors and demonstrates more robust performance across study areas.This is especially more pronounced when using small training samples(n<100).Examination of key parameters and water depth profiles highlights FFCLM’s advantages in generalization and deep-water inversion.This study presents an efficient solution for small-sample bathymetric mapping in turbid coastal waters,utilizing spectral and physical information to overcome sample size limitations and enhancing satellite remote sensing capabilities for shallow water monitoring.展开更多
As a highly vascular eye part,the choroid is crucial in various eye disease diagnoses.However,limited research has focused on the inner structure of the choroid due to the challenges in obtaining sufficient accurate l...As a highly vascular eye part,the choroid is crucial in various eye disease diagnoses.However,limited research has focused on the inner structure of the choroid due to the challenges in obtaining sufficient accurate label data,particularly for the choroidal vessels.Meanwhile,the existing direct choroidal vessel segmentation methods for the intelligent diagnosis of vascular assisted ophthalmic diseases are still unsatisfactory due to noise data,while the synergistic segmentation methods compromise vessel segmentation performance for the choroid layer segmentation tasks.Common cascaded structures grapple with error propagation during training.To address these challenges,we propose a cascade learning segmentation method for the inner vessel structures of the choroid in this paper.Specifically,we propose TransformerAssisted Cascade Learning Network(TACLNet)for choroidal vessel segmentation,which comprises a two-stage training strategy:pre-training for choroid layer segmentation and joint training for choroid layer and choroidal vessel segmentation.We also enhance the skip connection structures by introducing a multi-scale subtraction connection module designated as MSC,capturing differential and detailed information simultaneously.Additionally,we implement an auxiliary Transformer branch named ATB to integrate global features into the segmentation process.Experimental results exhibit that our method achieves the state-of-the-art performance for choroidal vessel segmentation.Besides,we further validate the significant superiority of the proposed method for retinal fluid segmentation in optical coherence tomography(OCT)scans on a publicly available dataset.All these fully prove that our TACLNet contributes to the advancement of choroidal vessel segmentation and is of great significance for ophthalmic research and clinical application.展开更多
Short-term photovoltaic(PV)power forecasting plays a crucial role in enhancing the stability and reliability of power grid scheduling.To address the challenges posed by complex environmental variables and difficulties...Short-term photovoltaic(PV)power forecasting plays a crucial role in enhancing the stability and reliability of power grid scheduling.To address the challenges posed by complex environmental variables and difficulties in modeling temporal features in PV power prediction,a short-term PV power forecasting method based on an improved CNN-LSTM and cascade learning strategy is proposed.First,Pearson correlation coefficients and mutual information are used to select representative features,reducing the impact of redundant features onmodel performance.Then,the CNN-LSTM network is designed to extract local features using CNN and learn temporal dependencies through LSTM,thereby obtaining feature representations rich in temporal information.Subsequently,a multi-layer cascade structure is developed,progressively integrating prediction results from base learners such as LightGBM,XGBoost,Random Forest(RF),and Extreme Random Forest(ERF)to enhance model performance.Finally,an XGBoost-based meta-learner is utilized to integrate the outputs of the base learners and generate the final prediction results.The entire cascading process adopts a dynamic expansion strategy,where the decision to add new cascade layers is based on the R2 performance criterion.Experimental results demonstrate that the proposed model achieves high prediction accuracy and robustness under various weather conditions,showing significant improvements over traditional models and providing an effective solution for short-term PV power forecasting.展开更多
Capturing high-fidelity normals from single face images plays a core role in numerous computer vision and graphics applications.Though significant progress has been made in recent years,how to effectively and efficien...Capturing high-fidelity normals from single face images plays a core role in numerous computer vision and graphics applications.Though significant progress has been made in recent years,how to effectively and efficiently explore normal priors remains challenging.Most existing approaches depend on the development of intricate network architectures and complex calculations for in-the-wild face images.To overcome the above issue,we propose a simple yet effective cascaded neural network,called Cas-FNE,which progressively boosts the quality of predicted normals with marginal model parameters and computational cost.Meanwhile,it can mitigate the imbalance issue between training data and real-world face images due to the progressive refinement mechanism,and thus boost the generalization ability of the model.Specifically,in the training phase,our model relies solely on a small amount of labeled data.The earlier prediction serves as guidance for following refinement.In addition,our shared-parameter cascaded block employs a recurrent mechanism,allowing it to be applied multiple times for optimization without increasing network parameters.Quantitative and qualitative evaluations on benchmark datasets are conducted to show that our Cas-FNE can faithfully maintain facial details and reveal its superiority over state-of-the-artmethods.The code is available at https://github.com/AutoHDR/CasFNE.git.展开更多
The umami taste,often described as the fifth basic taste,plays a pivotal role in the culinary world,significantly contributing to the overall flavor profile and consumer satisfaction of food products.Precise predictio...The umami taste,often described as the fifth basic taste,plays a pivotal role in the culinary world,significantly contributing to the overall flavor profile and consumer satisfaction of food products.Precise prediction and enhancement of umami taste intensity using the identification of umami peptides can lead to groundbreaking advancements in the food industry.This study presents a novel approach to combine machine learning and molecular docking techniques.The machine learning algorithm is based on a cascade algorithm combining CatBoost and BERT models to predict the umami taste intensity of peptides accurately.Our research reveals that combinations of specific amino acids,such as aspartic acid with alanine or glycine and lysine with glycine or histidine,have been identified to enhance the umami taste in foods.The model is available as a user-friendly web server at https://taste.infochemistry.ru.This study contributes to the scientific understanding of taste perception and provides a valuable tool for the food industry to innovate and improve product quality by optimizing umami taste profiles.展开更多
基金supported by the 2023 Hainan Province“South China Sea New Star”Science and Technology Innovation Talent Platform Project(NHXXRCXM202316)in part by Hainan Natural Science Foundation of China(nos.424QN253 and 620RC602)+5 种基金by the National Natural Science Foundation of China(no.61966013)in part by the Teaching Reform Research Project,Hainan Normal University,hsjg2023-07in part by the National Natural Science Foundation of China under grant 61991454in part by the National Key Research and Development Program of China under grant 2023Y FC3107605in part by the Oceanic Interdisciplinary Program of Shanghai Jiao Tong University under grant SL2022ZD206in part by the Scientific Research Fund of Second Institute of Oceanography,MNR under grant SL2302.
文摘Obtaining accurate bathymetric maps is very valuable for marine environment monitoring,port planning,and so on.Accurately estimating water depth in turbid coastal waters using satellite remote sensing encounters challenges originating from low water transparency,but it is limited by the quantity,quality,and water quality of samples.This study introduces a fast feature cascade learning model(FFCLM)to enhance the accuracy of bathymetric inversion from multispectral satellite images,particularly when limited field samples are available.FFCLM leverages spectral bands and in situ data to derive effective inversion weights through feature concatenation and cascade fitting.Field experiments conducted at Nanshan Port and Rushikonda Beach gathered water depth,satellite,and in situ data.Comparative analysis with conventional machine learning algorithms,including support vector machine,random forest,and gradient boosting trees,indicates that FFCLM achieves lower errors and demonstrates more robust performance across study areas.This is especially more pronounced when using small training samples(n<100).Examination of key parameters and water depth profiles highlights FFCLM’s advantages in generalization and deep-water inversion.This study presents an efficient solution for small-sample bathymetric mapping in turbid coastal waters,utilizing spectral and physical information to overcome sample size limitations and enhancing satellite remote sensing capabilities for shallow water monitoring.
基金supported by the National Natural Science Foundation of China under Grant Nos.62301330 and 62101346the Guangdong Basic and Applied Basic Research Foundation under Grant Nos.20231121103807001,2022A1515110101the Guangdong Provincial Key Laboratory under Grant No.2023B1212060076.
文摘As a highly vascular eye part,the choroid is crucial in various eye disease diagnoses.However,limited research has focused on the inner structure of the choroid due to the challenges in obtaining sufficient accurate label data,particularly for the choroidal vessels.Meanwhile,the existing direct choroidal vessel segmentation methods for the intelligent diagnosis of vascular assisted ophthalmic diseases are still unsatisfactory due to noise data,while the synergistic segmentation methods compromise vessel segmentation performance for the choroid layer segmentation tasks.Common cascaded structures grapple with error propagation during training.To address these challenges,we propose a cascade learning segmentation method for the inner vessel structures of the choroid in this paper.Specifically,we propose TransformerAssisted Cascade Learning Network(TACLNet)for choroidal vessel segmentation,which comprises a two-stage training strategy:pre-training for choroid layer segmentation and joint training for choroid layer and choroidal vessel segmentation.We also enhance the skip connection structures by introducing a multi-scale subtraction connection module designated as MSC,capturing differential and detailed information simultaneously.Additionally,we implement an auxiliary Transformer branch named ATB to integrate global features into the segmentation process.Experimental results exhibit that our method achieves the state-of-the-art performance for choroidal vessel segmentation.Besides,we further validate the significant superiority of the proposed method for retinal fluid segmentation in optical coherence tomography(OCT)scans on a publicly available dataset.All these fully prove that our TACLNet contributes to the advancement of choroidal vessel segmentation and is of great significance for ophthalmic research and clinical application.
基金2023 Sustainable Development Science and Technology Innovation Action Plan Project of Chongming District Science and Technology Committee,Shanghai(CKST2023-01)Shanghai Science and Technology Commission Funded Project(19DZ2254800).
文摘Short-term photovoltaic(PV)power forecasting plays a crucial role in enhancing the stability and reliability of power grid scheduling.To address the challenges posed by complex environmental variables and difficulties in modeling temporal features in PV power prediction,a short-term PV power forecasting method based on an improved CNN-LSTM and cascade learning strategy is proposed.First,Pearson correlation coefficients and mutual information are used to select representative features,reducing the impact of redundant features onmodel performance.Then,the CNN-LSTM network is designed to extract local features using CNN and learn temporal dependencies through LSTM,thereby obtaining feature representations rich in temporal information.Subsequently,a multi-layer cascade structure is developed,progressively integrating prediction results from base learners such as LightGBM,XGBoost,Random Forest(RF),and Extreme Random Forest(ERF)to enhance model performance.Finally,an XGBoost-based meta-learner is utilized to integrate the outputs of the base learners and generate the final prediction results.The entire cascading process adopts a dynamic expansion strategy,where the decision to add new cascade layers is based on the R2 performance criterion.Experimental results demonstrate that the proposed model achieves high prediction accuracy and robustness under various weather conditions,showing significant improvements over traditional models and providing an effective solution for short-term PV power forecasting.
基金supported by the National Natural Science Foundation of China(62072327)。
文摘Capturing high-fidelity normals from single face images plays a core role in numerous computer vision and graphics applications.Though significant progress has been made in recent years,how to effectively and efficiently explore normal priors remains challenging.Most existing approaches depend on the development of intricate network architectures and complex calculations for in-the-wild face images.To overcome the above issue,we propose a simple yet effective cascaded neural network,called Cas-FNE,which progressively boosts the quality of predicted normals with marginal model parameters and computational cost.Meanwhile,it can mitigate the imbalance issue between training data and real-world face images due to the progressive refinement mechanism,and thus boost the generalization ability of the model.Specifically,in the training phase,our model relies solely on a small amount of labeled data.The earlier prediction serves as guidance for following refinement.In addition,our shared-parameter cascaded block employs a recurrent mechanism,allowing it to be applied multiple times for optimization without increasing network parameters.Quantitative and qualitative evaluations on benchmark datasets are conducted to show that our Cas-FNE can faithfully maintain facial details and reveal its superiority over state-of-the-artmethods.The code is available at https://github.com/AutoHDR/CasFNE.git.
基金the state assignment of Ministry of Science and Higher Education of the Russian Federation(project N◦075-15-2024-483)Blue Sky Research and Prioritet 2030 Programs are acknowledged for infrastructural supportfinancial support from the Government of the Russian Federation through the ITMO Fellowship and Professorship Program.
文摘The umami taste,often described as the fifth basic taste,plays a pivotal role in the culinary world,significantly contributing to the overall flavor profile and consumer satisfaction of food products.Precise prediction and enhancement of umami taste intensity using the identification of umami peptides can lead to groundbreaking advancements in the food industry.This study presents a novel approach to combine machine learning and molecular docking techniques.The machine learning algorithm is based on a cascade algorithm combining CatBoost and BERT models to predict the umami taste intensity of peptides accurately.Our research reveals that combinations of specific amino acids,such as aspartic acid with alanine or glycine and lysine with glycine or histidine,have been identified to enhance the umami taste in foods.The model is available as a user-friendly web server at https://taste.infochemistry.ru.This study contributes to the scientific understanding of taste perception and provides a valuable tool for the food industry to innovate and improve product quality by optimizing umami taste profiles.
