arXiv subjects

John E. Taylor

Publications and source records attributed to John E. Taylor.

6 recordsLinked to original sources

arXiv2026

MCANet: A Multi-Scale Class-Specific Attention Network for Multi-Label Post-Hurricane Damage Assessment Using UAV Imagery

Hurricanes cause widespread damage to buildings, roads, and other infrastructure, making timely post-disaster damage assessment critical for emergency response and recovery planning. Unmanned aerial vehicle (UAV) imagery provides high-resolution observations of affected areas, but post-hurricane scenes are difficult to classify because multiple damage categories often co-occur within the same image, appear at different spatial scales, and include visually similar severity levels as well as rare but operationally important classes. To address these challenges, this study presents MCANet, a multi-label classification framework for post-hurricane UAV damage assessment. MCANet integrates a Res2Net-based backbone for multi-scale representation with class-specific residual attention to aggregate spatial evidence separately for each damage category. Evaluation on the RescueNet dataset, which includes 4,494 UAV images collected after Hurricane Michael and annotated with 10 damage categories, shows that MCANet achieves the highest mean average precision (mAP) among the evaluated models, with an mAP of 91.37%. Compared with Vision Transformer (ViT-B/16), MCANet improves mAP by 1.90 percentage points while using approximately 50% fewer parameters and 52% fewer giga floating-point operations (GFLOPs). Class-specific residual attention contributed an mAP increase of 1.52 percentage points over class-agnostic attention (CA); the Res2Net backbone alone produced concentrated gains on scale-heterogeneous or underrepresented classes such as Road Blocked and Pool. The single-head configuration achieved the highest overall mAP, whereas the two-head configuration provided targeted gains for Road Blocked and Building Major Damage. MCANet supports rapid post-disaster image screening and provides structured outputs for downstream emergency-management workflows.

cs.CV↗

arXiv2026

Knowledge-Guided Vision-Language Inference for Image-Based Urban Flood Depth Estimation

Timely floodwater depth estimates support road accessibility assessment and emergency response during urban flooding. Supervised vision methods often require extensive labeled datasets, while recent foundation vision-language models (VLMs) offer flexible visual reasoning but can inconsistently yield large errors in metric depth estimation. This paper proposes FloodVision, a knowledge-guided framework for estimating flood depth from a single RGB image. FloodVision integrates a general-purpose VLM with FloodKG, a domain knowledge base encoding canonical object dimensions and component landmarks (e.g., wheel arch, curb top) to encourage reasoning at the component level rather than treating objects as wholes. This injects explicit geometric grounding without task-specific training. Evaluated on 654 crowdsourced MyCoast New York flood images with resident-reported depths as proxy labels, FloodVision reduces the mean absolute error from 15.62 cm to 8.75 cm and the median error from 14.35 cm to 7.75 cm, with lower error than the VLM-only baseline in 69.3% of cases. The paper also discusses current limitations and future integration into urban digital twin systems.

cs.CV↗

arXiv2025

Aligning load flexibility with emissions reduction: empirical insights from a multi-site study of cryptocurrency data centers

The power sector is responsible for 32 percent of global greenhouse gas emissions. Data centers and cryptocurrencies use significant amounts of electricity and contribute to these emissions. Demand-side flexibility of data centers is one possible approach for reducing greenhouse gas emissions from these industries. To explore this, we use novel data collected from the Bitcoin mining industry to investigate the impact of load flexibility on power system decarbonization. Employing engineered metrics to explore curtailment dynamics and emissions alignment, we provide the first empirical analysis of cryptocurrency data centers' capability for reducing greenhouse gas emissions in response to real-time grid signals. Our results highlight the importance of strategically aligning operational behaviors with emissions signals to maximize avoided emissions. These findings offer insights for policymakers and industry stakeholders to enhance load flexibility and meet climate goals in these otherwise energy intensive data centers.

stat.AP↗

arXiv2025

Multi-Label Classification Framework for Hurricane Damage Assessment

Hurricanes cause widespread destruction, resulting in diverse damage types and severities that require timely and accurate assessment for effective disaster response. While traditional single-label classification methods fall short of capturing the complexity of post-hurricane damage, this study introduces a novel multi-label classification framework for assessing damage using aerial imagery. The proposed approach integrates a feature extraction module based on ResNet and a class-specific attention mechanism to identify multiple damage types within a single image. Using the Rescuenet dataset from Hurricane Michael, the proposed method achieves a mean average precision of 90.23%, outperforming existing baseline methods. This framework enhances post-hurricane damage assessment, enabling more targeted and efficient disaster response and contributing to future strategies for disaster mitigation and resilience. This paper has been accepted at the ASCE International Conference on Computing in Civil Engineering (i3CE 2025), and the camera-ready version will appear in the official conference proceedings.

cs.CV↗

arXiv2024

Short-term Streamflow and Flood Forecasting based on Graph Convolutional Recurrent Neural Network and Residual Error Learning

Accurate short-term streamflow and flood forecasting are critical for mitigating river flood impacts, especially given the increasing climate variability. Machine learning-based streamflow forecasting relies on large streamflow datasets derived from rating curves. Uncertainties in rating curve modeling could introduce errors to the streamflow data and affect the forecasting accuracy. This study proposes a streamflow forecasting method that addresses these data errors, enhancing the accuracy of river flood forecasting and flood modeling, thereby reducing flood-related risk. A convolutional recurrent neural network is used to capture spatiotemporal patterns, coupled with residual error learning and forecasting. The neural network outperforms commonly used forecasting models over 1-6 hours of forecasting horizons, and the residual error learners can further correct the residual errors. This provides a more reliable tool for river flood forecasting and climate adaptation in this critical 1-6 hour time window for flood risk mitigation efforts.

cs.AI↗

arXiv2014

Quantifying Human Mobility Perturbation and Resilience in Natural Disasters

Human mobility is influenced by environmental change and natural disasters. Researchers have used trip distance distribution, radius of gyration of movements, and individuals' visited locations to understand and capture human mobility patterns and trajectories. However, our knowledge of human movements during natural disasters is limited owing to both a lack of empirical data and the low precision of available data. Here, we studied human mobility using high-resolution movement data from individuals in New York City during and for several days after Hurricane Sandy in 2012. We found the human movements followed truncated power-law distributions during and after Hurricane Sandy, although the β value was noticeably larger during the first 24 hours after the storm struck. Also, we examined two parameters: the center of mass and the radius of gyration of each individual's movements. We found that their values during perturbation states and steady states are highly correlated, suggesting human mobility data obtained in steady states can possibly predict the perturbation state. Our results demonstrate that human movement trajectories experienced significant perturbations during hurricanes, but also exhibited high resilience. We expect the study will stimulate future research on the perturbation and inherent resilience of human mobility under the influence of natural disasters. For example, mobility patterns in coastal urban areas could be examined as tropical cyclones approach, gain or dissipate in strength, and as the path of the storm changes. Understanding nuances of human mobility under the influence of disasters will enable more effective evacuation, emergency response planning and development of strategies and policies to reduce fatality, injury, and economic loss.

physics.soc-ph↗