Article
Geochemistry & Geophysics
Quentin Brissaud, Elvira Astafyeva
Summary: This study presents a machine-learning-based framework to automatically detect co-seismic ionospheric disturbances (CIDs) and provide the possibility for real-time imaging of surface displacements. The machine learning models show excellent performance in classifying and picking CID arrival times compared to existing procedures.
GEOPHYSICAL JOURNAL INTERNATIONAL
(2022)
Article
Environmental Sciences
Maele Brisset, Simon Van Wynsberge, Serge Andrefouet, Claude Payri, Benoit Soulard, Emmanuel Bourassin, Romain Le Gendre, Emmanuel Coutures
Summary: Remote sensing, despite facing trade-offs between spatial and temporal resolution, proves to be an effective approach to monitor and backtrack the dynamics of green algae blooms in shallow lagoons using Sentinel-2 satellite data. Through comparing spectral indices with field observations, significant insights into the variations of algal blooms can be achieved.
Article
Geosciences, Multidisciplinary
Catherine V. L. Pennington, Remy Bossu, Ferda Ofli, Muhammad Imran, Umair Qazi, Julien Roch, Vanessa J. Banks
Summary: This paper describes and validates the development of a system that continuously monitors social media for landslide-related content and identifies the most relevant information using a landslide classification model. The system has been quantitatively verified to detect landslide reports with a precision of 76% during real-world deployment. The next stage of development will incorporate stakeholder and user feedback.
INTERNATIONAL JOURNAL OF DISASTER RISK REDUCTION
(2022)
Article
Ecology
P. Kamermans, A. J. Murk, T. Wijgerde
Summary: Coral and oyster reefs have suffered significant declines due to human activities, resulting in low recruitment rates of larvae and hindering ecosystem recovery. This study introduces a novel approach using high resolution tracking technology to analyze the swimming and substrate selection behavior of coral and oyster larvae in great detail. By analyzing second-by-second spatial data, the researchers were able to extract variables such as swimming pattern, speed, and distance traveled, providing insights into the behavior of these larvae. This study has potential applications in ecology, aquaculture, and coastal engineering, particularly in developing substrates to promote larval settlement on reefs for restoration efforts.
JOURNAL OF EXPERIMENTAL MARINE BIOLOGY AND ECOLOGY
(2023)
Article
Environmental Sciences
Yip Hung Yeung, James Y. Xie, Chun Kit Kwok, Keith Kei, Put Ang, Leo Lai Chan, Walter Dellisanti, Chi Chiu Cheang, Wing Kuen Chow, Jian-Wen Qiu
Summary: The study identified five community types in Hong Kong's coral communities, with the most common one dominated by massive and upward-plating corals. Coral cover and generic richness were negatively correlated with water quality parameters, indicating constraints on the development of coral communities. Management actions are recommended to reduce bioerosion and monitor sites affected by bleaching.
MARINE POLLUTION BULLETIN
(2021)
Article
Chemistry, Multidisciplinary
Francesco Carrera, Vincenzo Dentamaro, Stefano Galantucci, Andrea Iannacone, Donato Impedovo, Giuseppe Pirlo
Summary: The 0-day attack is a cyber-attack that exploits unpublished vulnerabilities. Detecting and predicting such attacks is crucial for smart enterprises and technology-dependent systems. Unsupervised machine learning methods are effective in identifying anomalies in real-time. The addition of Isolation Forest improves accuracy and inference time. The study also uses SHAP to identify important features for classifying attack events. Experiments were conducted on multiple datasets.
APPLIED SCIENCES-BASEL
(2022)
Article
Materials Science, Multidisciplinary
Byungjoon Bae, Yongmin Baek, Jeongyong Yang, Heesung Lee, Charana S. S. Sonnadara, Sangeun Jung, Minseong Park, Doeon Lee, Sihwan Kim, Gaurav Giri, Sahil Shah, Geonwook Yoo, William A. A. Petri, Kyusang Lee
Summary: In order to achieve accurate diagnosis and immunity to viruses, a IGZO-based biosensor field-effect transistor has been developed which can simultaneously detect viral antigens and corresponding antibodies in less than 20 minutes with high accuracy. This system will play a crucial role in preventing global viral outbreaks.
Article
Robotics
Haitao Meng, Changcai Li, Chonghao Zhong, Jianfeng Gu, Gang Chen, Alois Knoll
Summary: This paper presents FastFusion, a three-stage stereo-LiDAR deep fusion scheme that integrates LiDAR information into each step of classical stereo-matching taxonomy, achieving real-time high-precision dense depth sensing. By utilizing a compact binary neural network, the stereo-LiDAR information is integrated, and a proposed cross-based LiDAR trust aggregation further fuses the sparse LiDAR measurements in the back-end of stereo matching. A refinement network is introduced to ensure consistency between the photometric of the input image and the depth estimation. Additionally, a GPU-based acceleration framework is proposed to provide low-latency implementation of FastFusion, improving both accuracy and real-time responsiveness.
JOURNAL OF FIELD ROBOTICS
(2023)
Review
Psychology, Multidisciplinary
Muhammad Usman Tariq, Marc Poulin, Abdullah A. Abonamah
Summary: This paper provides a detailed literature review on the driving forces and barriers for achieving operational excellence through artificial intelligence (AI). AI is a broad technological concept that includes operational management, philosophy, humanities, statistics, mathematics, computer sciences, and social sciences. Developments in technology and advancements in producing intelligence for machines have a positive impact on decision-making, operations, strategies, and management in the production process of goods and services.
FRONTIERS IN PSYCHOLOGY
(2021)
Article
Multidisciplinary Sciences
Ahmed Abdelaal, Salaheldin Elkatatny, Abdulazeez Abdulraheem
Summary: Two models were developed using artificial neural networks and adaptive neuro-fuzzy inference system to estimate formation pressure gradient in real-time through drilling data. The models showed good accuracy in predicting pressure gradient, proving their reliability.
SCIENTIFIC REPORTS
(2022)
Article
Green & Sustainable Science & Technology
Beilin Liu, Zhiqiang Liu, Jingzheng Ren, Nan Xie, Sheng Yang
Summary: This study proposes a scheduling optimization model based on a mixed integer nonlinear programming model for multi-energy complementary integrated energy systems (MCIES) and applies it to a swimming pool in Changsha. The results show that real-time operational scheduling considering flexibility can enhance the system's flexibility and reduce carbon emissions.
JOURNAL OF CLEANER PRODUCTION
(2023)
Article
Chemistry, Multidisciplinary
Anderson B. B. Mayfield, Alexandra C. C. Dempsey, Chii-Shiarng Chen, Chiahsin Lin
Summary: Current models for predicting coral health often focus only on temperature and coral abundance, neglecting other influential factors. To develop more reliable predictions, researchers trained an artificial intelligence to predict coral stress susceptibility using seawater quality, benthic survey, and molecular biomarker data. The neural network model can accurately predict coral health using cheaper and easier-to-measure environmental and ecological features.
APPLIED SCIENCES-BASEL
(2022)
Article
Environmental Sciences
Natalie Levy, Noa Simon-Blecher, Shachaf Ben-Ezra, Matan Yuval, Tirza Doniger, Matthieu Leray, Sarit Karako-Lampert, Ezri Tarazi, Oren Levy
Summary: This study demonstrates the use of eDNA metabarcoding as a tool for quantifying coral reef biodiversity, particularly for cryptofauna and organisms in early life stages. The results show that eDNA metabarcoding is effective in comprehensively evaluating invertebrate communities on complex 3D structures and understanding the role of these structures in providing habitat for organisms.
SCIENCE OF THE TOTAL ENVIRONMENT
(2023)
Article
Geography, Physical
Xuewen Wang, Xianmin Wang, Xinlong Zhang, Lizhe Wang, Haixiang Guo, Dongdong Li
Summary: This paper proposes a novel interpretable self-supervised learning (ISeL) method for the near real-time spatial prediction of earthquake-induced landslides (EQILs). The method introduces swap noise and an interpretable module to improve the generalization and transferability of the model, and reduce false alarm and improve accuracy. Experimental results demonstrate the superiority of the ISeL model over state-of-the-art machine learning and deep learning methods, and its potential application in earthquake-frequent regions worldwide.
INTERNATIONAL JOURNAL OF DIGITAL EARTH
(2023)
Article
Chemistry, Multidisciplinary
Anderson B. Mayfield, Chiahsin Lin
Summary: An artificial intelligence trained with protein concentration data accurately predicted bleaching susceptibility of massive corals in the Upper Florida Keys. Proteomic data from laboratory and field samples were used to train neural networks and machine-learning models, which were then tested with massive corals sampled during a bleaching event in 2019. The resulting AI was capable of accurately predicting coral bleaching based solely on protein signatures, suggesting potential applications for assessing climate resilience in Orbicella faveolata.
APPLIED SCIENCES-BASEL
(2023)