检索结果-内蒙古大学图书馆

DiffEGG: Diffusion-Driven Edge Generation as a Pixel-Annotation-Free Alternative for Instance Annotation

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Jo, Sanghyun Lee, Ziseok Lee, Wooyeol Kim, Kyungsu OGQ Seoul Korea Republic of Department of Biomedical Sciences Seoul National University Seoul Korea Republic of Department of Computer Science and Engineering Seoul National University Seoul Korea Republic of School of Transdisciplinary Innovations Interdisciplinary Program in Artificial Intelligence Seoul National University Seoul Korea Republic of

Achieving precise panoptic segmentation heavily relies on pixel-wise instance annotations, but obtaining such datasets is costly. Unsupervised instance segmentation (UIS) eliminates annotation requirements but struggles with adjacent instance merging and single-instance fragmentation, largely due to the limitations of DINO-based backbones, which lack strong instance separation cues. Weakly-supervised panoptic segmentation (WPS) reduces annotation costs by using sparse labels (e.g., points, boxes), yet these annotations remain expensive and introduce human bias and boundary errors, limiting performance. To address these challenges, we propose DiffEGG (Diffusion-Driven EdGe Generation), a fully annotation-free method that extracts instance-aware features from pretrained diffusion models to generate precise instance edge maps. DiffEGG fundamentally differs from prior UIS and WPS approaches by leveraging pretrained diffusion models as a backbone. Unlike DINO-based UIS methods, diffusion models inherently capture fine-grained, instance-aware features, enabling more precise boundary delineation. DiffEGG further enhances edge connectivity through self-supervised fine-tuning, progressively improving instance separation. For WPS, DiffEGG completely eliminates annotation costs and human bias by operating without any form of manual supervision, addressing the key limitations of prior best methods. Additionally, we introduce RIP, a post-processing technique that fuses DiffEGG’s edge maps with segmentation masks in a task-agnostic manner. This broad applicability allows DiffEGG to be seamlessly integrated into various segmentation frameworks. When applied to UIS, DiffEGG and RIP achieve an average +4.4 AP improvement over prior best UIS methods. When combined with weakly-supervised semantic segmentation (WSS), DiffEGG enables WPS without instance annotations, outperforming prior best point-supervised WPS methods by +1.7 PQ. These results demonstrate that DiffEGG’s edge maps serv

关键词： Semantic Segmentation

Social network analysis of manga: similarities to real-world social networks and trends over decades

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Sugishita, Kashin Masuda, Naoki Department of Transdisciplinary Science and Engineering Tokyo Institute of Technology Tokyo152-8550 Japan Department of Mathematics State University of New York at Buffalo BuffaloNY14260-2900 United States Computational and Data-Enabled Science and Engineering Program State University of New York at Buffalo BuffaloNY14260-5030 United States Center for Computational Social Science Kobe University Kobe657-8501 Japan

Manga, Japanese comics, has been popular on a global scale. Social networks among characters, which are often called character networks, may be a significant contributor to their popularity. We collected data from 162 popular manga that span over 70 years and analyzed their character networks. First, we found that many of static and temporal properties of the character networks are similar to those of real human social networks. Second, the character networks of most manga are protagonist-centered such that a single protagonist interacts with the majority of other characters. Third, the character networks for manga mainly targeting boys have shifted to denser and less protagonist-centered networks and with fewer characters over decades. Manga mainly targeting girls showed the opposite trend except for the downward trend in the number of characters. The present study, which relies on manga data sampled on an unprecedented scale, paves the way for further population studies of character networks and other aspects of comics. Copyright © 2023, The Authors. All rights reserved.

关键词： Population statistics

Spatio-Temporal Evaluation of Urban Growth and Surface Urban Heat Island from 2000 to 2019 in Java Island, Indonesia

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Fajary, Faiz Rohman Lee, Han Soo Kubota, Tetsu Bhanage, Vinayak Pradana, Radyan Putra Nimiya, Hideyo Dewa Gede, I. Arya Putra Transdisciplinary Science and Engineering Program Graduate School of Advanced Science and Engineering Hiroshima University 1-5-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8529 Japan Atmospheric Science Research Group Faculty of Earth Science and Technology Institut Teknologi Bandung West Java Bandung40132 Indonesia The IDEC Institute Hiroshima University 1-5-1 Kagamiyama Higashi-Hiroshima Hiroshima739-8529 Japan Graduate School of Science and Engineering Kagoshima University 1-21-40 Korimoto Kagoshima890-0065 Japan Jl. Angkasa 1 No. 2 Kec. Kemayoran Jakarta10610 Indonesia

This study aims to classify land use and land cover (LULC) and analyze urban growth and its effects on urban heat island using remote sensing in the western part of Java Island and the three focused urban areas: Jakarta metropolitan area (JMA), Bandung and Cimahi municipalities (BC), and Sukabumi municipality (SKB). Landsat imagery with 30 m spatial resolution from three years was used: 2000, 2009, and 2019. For LULC classification, the support vector machine method delivers good accuracy. The city clustering algorithm was adopted to define the urban area for surface urban heat island intensity (SUHII) calculation. In the last two decades, urban areas have grown by more than twice in JMA and SKB and nearly 1.5 times in BC. Along with the urban growth, SUHII in JMA increases from 1.94 to 3.71 ºC, similar to BC but with a lower magnitude. Meanwhile, it fluctuates between 3.05 and 3.56 ºC in SKB. For the unchanged urban pixels, the LST exhibits positive linear trends with magnitudes of 0.25, 0.16, and 0.14 ºC/year in JMA, SKB, and BC, respectively. The findings of this study will be valuable to the government in evaluating and planning sustainable cities. © 2023, The Authors. All rights reserved.

关键词： Urban growth

Ocean thermal energy conversion (OTEC) potential in central American and Caribbean regions: A multicriteria analysis for optimal sites

学校读者我要写书评

暂无评论

Applied Energy 2025年 394卷

作者： Soto Calvo, Manuel Lee, Han Soo Coastal Hazards and Energy System Science Laboratory Transdisciplinary Science and Engineering Program of Advanced Science and Engineering Hiroshima University 1-5-1 Kagamiyama Hiroshima Higashi-Hiroshima739-8529 Japan The IDEC institute Hiroshima University Japan Smart Energy of Innovation and Practice for Smart Society Hiroshima University Japan

This study presents a comprehensive basin-wide assessment of ocean thermal energy conversion (OTEC) potential in Central America and the Caribbean region, integrating high-resolution oceanographic data from the Copernicus Marine Service with advanced technoeconomic modelling. The analysis employs a multicriteria approach combining thermal resource assessment, technical feasibility evaluation, and detailed economic analysis across the entire Caribbean Basin. The results identify several high-potential areas, particularly along Mexico's Caribbean coast, the Greater Antilles (notably Cuba's southeastern coast), and the Lesser Antilles chain, with temperature differentials reaching over 25 °C in high-potential areas and maintaining levels consistently above 24 °C year-round. For utility-scale OTEC plants, the analysis reveals potential annual electricity generation ranging from 400 to 933 GWh/year per territory, with optimal levelized cost of energy (LCOE) values between 120 and 350 $/MWh at the most favourable locations and net power efficiencies of 2.5–3 %. The study quantifies key technical parameters and identifies five distinct regional clusters with varying characteristics, suggesting the need for tailored implementation strategies. Economic analysis indicates that heat exchangers and cold-water pipes represent approximately 55.4 % of the total CAPEX, whereas Monte Carlo risk assessment reveals that project viability is most sensitive to capital cost variations and capacity factor uncertainties. The research demonstrated high plant availability, with capacity factors exceeding 85 % in optimal locations, suggesting that OTEC could serve as a viable baseload renewable energy option for many Caribbean territories. This comprehensive assessment provides a robust framework for evaluating OTEC potential in tropical maritime regions while offering actionable insights for policymakers and energy planners considering OTEC as part of their renewable energy portfolio. © 2025

关键词： Heat generation

Integrated tidal power potential in the Seto Inland Sea in Japan with cost–benefit analysis

学校读者我要写书评

暂无评论

Applied Energy 2025年 395卷

作者： Morhaf Aljber Han Soo Lee Jae-Soon Jeong Coastal Hazards and Energy System Science Laboratory Graduate School of Innovation and Practice for Smart Society Hiroshima University 1-5-1 Kagamiyama Higashi-Hiroshima 739-8529 Hiroshima Japan Center for Planetary Health and Innovation Science (PHIS) IDEC Institute Hiroshima University Japan Transdisciplinary Science and Engineering Program Graduate School of Advanced Science and Engineering Hiroshima University Japan

Tidal power generation (TPG) provides a predictable and stable energy supply, making it a valuable complement to variable renewable sources such as solar and wind. However, its commercial viability remains constrained by high capital costs, complex installation requirements, and limited specialised infrastructure. This study assesses the economic feasibility of TPG in the Seto Inland Sea in Japan using the levelised cost of energy (LCOE) as a benchmark for competitiveness. The findings indicated that with a 10 % discount rate, the LCOE ave is approximately 85,990 ¥/MWh, whereas a 5 % discount rate reduces it to 31,088 ¥/MWh. The estimated annual energy production (AEP) was 374,640 MWh, sufficient to supply 89,735 households, with an equivalent CO₂ reduction of approximately 166,000 tons per year. Achieving cost competitiveness for TPGs requires large-scale deployment and supportive government incentives. Expanding infrastructure and leveraging economies of volume are crucial for reducing costs and enabling broader adoption. Despite existing challenges, TPG holds significant potential for enhancing Japan's energy security and contributing to its decarbonisation goals. This potential depends on effective grid integration, strong policy support, and sustained strategic investment.

关键词： Tidal power competitiveness Levelised cost of energy Marine renewable price Sustainable energy transition LCOE Carbone neutrality

A growth model for water distribution networks with loops

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Sugishita, Kashin Abdel-Mottaleb, Noha Zhang, Qiong Masuda, Naoki Department of Mathematics State University of New York at Buffalo BuffaloNY14260-2900 United States Department of Transdisciplinary Science and Engineering Tokyo Institute of Technology Tokyo152-8550 Japan Department of Civil and Environmental Engineering University of South Florida TampaFL33620 United States Computational and Data-Enabled Science and Engineering Program State University of New York at Buffalo BuffaloNY14260-5030 United States Faculty of Science and Engineering Waseda University Tokyo169-8555 Japan

Water distribution networks (WDNs) expand their service areas over time. These growth dynamics are poorly understood. One facet of WDNs is that they have loops in general, and closing loops may be a functionally important process for enhancing their robustness and efficiency. We propose a growth model for WDNs which generates networks with loops and is applicable to networks with multiple water sources. We apply the proposed model to four empirical WDNs to show that it produces networks whose structure is similar to that of the empirical WDNs. The comparison between the empirical and modeled WDNs suggests that the empirical WDNs may realize a reasonable balance between cost, efficiency, and robustness in terms of the network structure. We also study the design of pipe diameters based on a biological positive feedback mechanism. Specifically, we apply a model inspired by Physarum polycephalum to find moderate positive correlations between the empirical and modeled pipe diameters. The difference between the empirical and modeled pipe diameters suggests that we may be able to improve the performance of WDNs by following organizing principles of biological flow networks. Copyright © 2021, The Authors. All rights reserved.

关键词： Water distribution systems

Discovery of two infrared objects with strong ice absorption in the AKARI slit-less spectroscopic survey of the Galactic Plane

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Onaka, Takashi Kimura, Tomoyuki Sakon, Itsuki Shimonishi, Takashi Department of Physics Faculty of Science and Engineering Meisei University 2-1-1 Hodokubo Hino Tokyo191-8506 Japan Department of Astronomy Graduate School of Science The University of Tokyo 7-3-1 Hongo Bunkyo-ku Tokyo113-0033 Japan Center for Transdisciplinary Research Niigata University Ikarashi-nihoncho 8050 Nishi-ku Niigata950-2181 Japan Environmental Science Program Department of Science Faculty of Science Niigata University Ikarashi-nihoncho 8050 Nishi-ku Niigata950-2181 Japan

We discover two infrared objects that show deep absorption features of H2O, CO2, and CO ices in the AKARI/Infrared Camera (IRC) slit-less spectroscopic survey of the Galactic plane in 2.5-13 µm. Both objects are located neither in known star-forming regions nor in known dense clouds. For one of the objects, Object 1, we successfully extract a spectrum from 2.5 to 13 µm, which also shows several absorption features in 5-13 µm, including deep silicate absorption at 10 µm. For the other object, Object 2, only a spectrum from 3.1 to 5 µm is reliably extracted due to the presence of nearby overlapping objects and faint nebulosity. Both objects show warm (> 100 K) CO gas absorption in addition to the ice absorption features, suggesting that they are embedded young stellar objects (YSOs). On the other hand, both objects have spectral energy distributions (SEDs) that peak at around 5 µm and decrease towards longer wavelengths. These characteristics of the SEDs and the presence of deep absorption features cannot easily be accounted for by standard YSO models. They may be explained as background stars behind dense clouds. We discuss possible nature of the objects and implications of the present discovery. © 2021, CC BY-NC-ND.

关键词： Ice

Chemical evolution of some selected complex organic molecules in low-mass star-forming regions

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Bhat, Bratati Kar, Rumela Mondal, Suman Kumar Ghosh, Rana Gorai, Prasanta Shimonishi, Takashi Tanaka, Kei E.I. Furuya, Kenji Das, Ankan Institute of Astronomy Space and Earth Science 1 Bidhan Sishu Sarani Kolkata700054 India Department of Metallurgical Engineering and Material Sciences IIT Bombay Powai Maharashtra Mumbai400076 India S. N. Bose National Center for Basic Sciences JD-Block Salt Lake Kolkata700098 India Indian Centre for Space Physics 43 Chalantika Garia Station Road Kolkata700084 India Department of Space Earth & Environment Chalmers University of Technology GothenburgSE-412 96 Sweden Center for Transdisciplinary Research Niigata University Niigata Nishi-ku950-2181 Japan Environmental Science Program Department of Science Faculty of Science Niigata University Niigata Nishi-ku950-2181 Japan Department of Earth and Planetary Sciences Tokyo Institute of Technology Meguro Tokyo152-8551 Japan Center for Computational Sciences University of Tsukuba Tsukuba305-8577 Japan National Astronomical Observatory of Japan Tokyo181-8588 Japan

The destiny of complex organic molecules (COMs) in star-forming regions is interlinked with various evolutionary phases. Therefore, identifying these species in diversified environments of identical star-forming regions would help to comprehend their physical and chemical heritage. We identified multiple COMs utilizing the Large program ‘Astrochemical Surveys At IRAM’ (ASAI) data, dedicated to chemical surveys in Sun-like star-forming regions with the IRAM 30 m telescope. It was an unbiased survey in the millimeter regime, covering the prestellar core, protostar, outflow region, and protoplanetary disk phase. Here, we have reported some transitions of seven COMs, namely, methanol (CH3OH), acetaldehyde (CH3CHO), methyl formate (CH3OCHO), ethanol (C2H5OH), propynal (HCCCHO), dimethyl ether (CH3OCH3), and methyl cyanide (CH3CN) in some sources L1544, B1-b, IRAS4A, and SVS13A. We found a trend among these species from the derived abundances using the rotational diagram method and MCMC fit. We have found that the abundances of all of the COMs, except for HCCCHO, increase from the L1544 (prestellar core) and peaks at IRAS16293-2422 (class 0 phase). It is noticed that the abundance of these molecules correlate with the luminosity of the sources. The obtained trend is also visible from the previous interferometric observations and considering the beam dilution effect. Copyright © 2023, The Authors. All rights reserved.

关键词： Molecules

MedRep: Medical Concept Representation for General Electronic Health Record Foundation Models

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Kim, Junmo Lee, Namkyeong Kim, Jiwon Kim, Kwangsoo Interdisciplinary Program in Bioengineering Seoul National University Korea Republic of Dept. of Industrial and Systems Engineering KAIST Korea Republic of Interdisciplinary Program of Medical Informatics Seoul National University Korea Republic of Dept. of Transdisciplinary Medicine ICMIT Seoul National University Hospital Korea Republic of Center for Data Science Healthcare AI Research Institute Seoul National University Hospital Korea Republic of Dept. of Medicine College of Medicine Seoul National University Korea Republic of

Electronic health record (EHR) foundation models have been an area ripe for exploration with their improved performance in various medical tasks. Despite the rapid advances, there exists a fundamental limitation: Processing unseen medical codes out of the vocabulary. This problem limits the generality of EHR foundation models and the integration of models trained with different vocabularies. To deal with this problem, we propose MedRep for EHR foundation models based on the observational medical outcome partnership (OMOP) common data model (CDM), providing the integrated medical concept representations and the basic data augmentation strategy for patient trajectories. For concept representation learning, we enrich the information of each concept with a minimal definition through large language model (LLM) prompts and enhance the text-based representations through graph ontology of OMOP vocabulary. Trajectory augmentation randomly replaces selected concepts with other similar concepts that have closely related representations to let the model practice with the concepts out-of-vocabulary. Finally, we demonstrate that EHR foundation models trained with MedRep better maintain the prediction performance in external datasets. Our code implementation is publicly available at https://***/kicarussays/MedRep. Copyright © 2025, The Authors. All rights reserved.

关键词： Electronic health record