This article describes the development and implementations of a novel software platform that supports real-time, science-based policy making on air quality through a user-friendly interface. The software, RSM-VAT, use...This article describes the development and implementations of a novel software platform that supports real-time, science-based policy making on air quality through a user-friendly interface. The software, RSM-VAT, uses a response surface modeling（RSM） methodology and serves as a visualization and analysis tool（VAT） for three-dimensional air quality data obtained by atmospheric models. The software features a number of powerful and intuitive data visualization functions for illustrating the complex nonlinear relationship between emission reductions and air quality benefits. The case study of contiguous U.S.demonstrates that the enhanced RSM-VAT is capable of reproducing the air quality model results with Normalized Mean Bias 〈 2% and assisting in air quality policy making in near real time.展开更多
Prompt learning has become crucial for adapting Visual Language Models(VLM)to downstream tasks.Although existing prompt learning models have made significant strides,they still face two major challenges:1.Too much att...Prompt learning has become crucial for adapting Visual Language Models(VLM)to downstream tasks.Although existing prompt learning models have made significant strides,they still face two major challenges:1.Too much attention is paid to learning about basic classes,making it harder to understand novel classes;2.Most methods only rely on the context information provided by the prompt template,resulting in limited text features.In this study,we propose a new fine-tuning method for Visual-Language Models called Input-Enhanced Prompt Tuning(IEPT).The IEPT improves the generalization of VLMs for downstream tasks by introducing two components,i.e.,the Data Augmentation Framework(DAF)and the Category Generalization Optimizer(CGO).Specifically,the DAF employs Large Language Models to resolve issues of word ambiguity by obtaining more class label context,and uses simple image augmentation to address the issue of limited features by providing more image samples.The CGO prevents overfitting by adding new class names during training.Experiments show that the performance of IEPT in various evaluation suites is better or comparable to that of the existing method,covering basic to novel generalization,domain generalization,and cross-dataset evaluation.Compared to the state-of-the-art method PromptSRC,IEPT achieves an absolute improvement of 0.40%for base classes,1.56%for novel classes and 1.04%on the harmonic mean,averaged over 11 datasets.In addition,we present detailed ablation studies that validate the individual contributions of DAF and CGO to the overall performance of IEPT.Our code is available at https://github.com/ayuan 0626/IEPT.展开更多
基金Financial and data support for this work is provided by the U.S. Environmental Protection Agency (No. GS-10F-0205T)partly supported by the funding of Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control (No. h2xj D612004 Ш )+1 种基金the funding of State Environmental Protection Key Laboratory of Sources and Control of Air Pollution Complex (No. SCAPC201308)the project of Atmospheric Haze Collaboration Control Technology Design (No. XDB05030400) from Chinese Academy of Sciences
文摘This article describes the development and implementations of a novel software platform that supports real-time, science-based policy making on air quality through a user-friendly interface. The software, RSM-VAT, uses a response surface modeling（RSM） methodology and serves as a visualization and analysis tool（VAT） for three-dimensional air quality data obtained by atmospheric models. The software features a number of powerful and intuitive data visualization functions for illustrating the complex nonlinear relationship between emission reductions and air quality benefits. The case study of contiguous U.S.demonstrates that the enhanced RSM-VAT is capable of reproducing the air quality model results with Normalized Mean Bias 〈 2% and assisting in air quality policy making in near real time.
基金supported by National Key R&D Program of China(No.2022YFE0196100)the Innovation Capacity Enhancement Program Science and Technology Platform Project of Hebei Province(22567623H)Hebei University High Level Innovative Talent Research Start-up Funding Project(No.521000981092).
文摘Prompt learning has become crucial for adapting Visual Language Models(VLM)to downstream tasks.Although existing prompt learning models have made significant strides,they still face two major challenges:1.Too much attention is paid to learning about basic classes,making it harder to understand novel classes;2.Most methods only rely on the context information provided by the prompt template,resulting in limited text features.In this study,we propose a new fine-tuning method for Visual-Language Models called Input-Enhanced Prompt Tuning(IEPT).The IEPT improves the generalization of VLMs for downstream tasks by introducing two components,i.e.,the Data Augmentation Framework(DAF)and the Category Generalization Optimizer(CGO).Specifically,the DAF employs Large Language Models to resolve issues of word ambiguity by obtaining more class label context,and uses simple image augmentation to address the issue of limited features by providing more image samples.The CGO prevents overfitting by adding new class names during training.Experiments show that the performance of IEPT in various evaluation suites is better or comparable to that of the existing method,covering basic to novel generalization,domain generalization,and cross-dataset evaluation.Compared to the state-of-the-art method PromptSRC,IEPT achieves an absolute improvement of 0.40%for base classes,1.56%for novel classes and 1.04%on the harmonic mean,averaged over 11 datasets.In addition,we present detailed ablation studies that validate the individual contributions of DAF and CGO to the overall performance of IEPT.Our code is available at https://github.com/ayuan 0626/IEPT.
