Results 61 to 70 of about 9,649,421 (248)

ConWave-LoRA: Concept Fusion in Customized Diffusion Models with Contrastive Learning and Wavelet Filtering

Computers
Customizing diffusion models via Low-Rank Adaptation (LoRA) has become a standard approach for customized concept injection. However, synthesizing multiple customized concepts within a single image remains challenging due to the parameter pollution ...
Xinying Liu, Xiaogang Huo, Zhihui Yang
doaj   +1 more source

LoRA-XS: Low-Rank Adaptation with Extremely Small Number of Parameters

The growth of large language models underscores the need for parameter-efficient fine-tuning. Despite its popularity, LoRA encounters storage and computational challenges when deploying multiple task- or user-specific modules. To address this, we introduce LoRA-XS, a novel fine-tuning method backed by a theoretical derivation.
Bałazy, Klaudia, Banaei, Mohammadreza, Aberer, Karl, Tabor, Jacek   +3 more
openaire   +3 more sources

Resilience for restoration: Insights from the Northern Great Plains native seed system

People and Nature, EarlyView.
Abstract Native plants are essential to ecological restoration and culturally important to local and Tribal communities, yet native seed systems across the United States (US) struggle to meet restoration demand. In the US Northern Great Plains, land conversion of native ecosystems means restoration potential is high but constrained by an underdeveloped
Elena Friedman, Lauren Redmore, Lora Perkins, Jennifer Zavaleta Cheek   +3 more
wiley   +1 more source

Synthetic Imagery Improves Ecological Classification When Real Data Are Scarce, Not Direct Substitutes

Remote Sensing in Ecology and Conservation, EarlyView.
Image generation models can create synthetic ecological datasets and potentially alleviate limitations from sparse real data. Using 20 tree genera across community‐science and street‐view imagery, we tested whether synthetic images could replace or supplement real data for image classification.
Thomas A. Lake, Brennen T. Farrell, Ross K. Meentemeyer, Chris M. Jones   +3 more
wiley   +1 more source

Parameter-Efficient Fine-Tuning for Photovoltaic Cell Defect Classification: A Systematic Comparison of LoRA, QLoRA, and Full Fine-Tuning on ConvNeXt-Tiny

Sensors
Automated visual inspection of photovoltaic (PV) cells is an important component of solar-module quality assurance. However, adapting modern pre-trained vision backbones to PV defect classification remains challenging because full fine-tuning requires ...
Seda Bayat Toksöz, Gültekin Işık, Gökhan Şahin, Erdal Akin   +3 more
doaj   +1 more source

Artificial Intelligence for Blue Transformation: A Review of Predictive Modeling and Decision Support Systems in Sustainable Aquaculture

Sustainable Development, EarlyView.
ABSTRACT The FAO's “blue transformation” roadmap necessitates a fundamental shift towards precision aquaculture to meet global food security targets while minimizing environmental footprints. This review provides a comprehensive overview of how artificial intelligence (AI) and decision support systems (DSS) serve as pivotal enablers for the “better ...
Mustafa Öz, Enes Üstüner
wiley   +1 more source

Fine-Tuning Open-Source Large Language Models for Cybersecurity and IT Support: A Comparative Study on LLM Fine-Tuning Techniques

IEEE Open Journal of the Computer Society
The focus of Artificial Intelligence (AI) has progressed from exploring whether Large Language Models (LLMs) can transform domain-specific tasks to understanding how to optimise their impact.
Chaithanya Vamshi Sai, Nouh Sabri Elmitwally, Iain Rice, Haitham Mahmoud, Ian Vickers, Xavier Schmoor   +5 more
doaj   +1 more source

Operationalizing the Waste Hierarchy Though A Synthetic Performance Indicator: Evidence From Spain

Sustainable Development, EarlyView.
ABSTRACT Ensuring compliance with the European Union (EU) waste hierarchy is essential for advancing circular economy objectives and meeting legally binding municipal solid waste (MSW) targets. This study develops and applies the Waste Hierarchy Achievement Synthetic Indicator (WHASI), an innovative synthetic index designed to assess territorial ...
Victor Mocholí, Trinidad Gómez, María Molinos‐Senante   +2 more
wiley   +1 more source

Parameter-efficient fine-tuning of SAM for SAR flood mapping: comparative performance and encoder–decoder representation analysis

International Journal of Digital Earth
Synthetic aperture radar (SAR) enables flood mapping under cloud and adverse weather, but segmentation remains challenging across heterogeneous environments.
Ziming Wang, Yigong Hu, Jeffrey Neal, Peter M. Atkinson, Ce Zhang   +4 more
doaj   +1 more source

DQN‐Guided Subset‐Induced OCSVM Kernel Approximation for Imbalanced Anomaly Detection

IEEJ Transactions on Electrical and Electronic Engineering, EarlyView.
Anomaly detection under limited normal data remains a fundamental challenge due to severe class imbalance and scarcity of anomalies. We propose a novel framework that reformulates support vector selection in One‐Class SVM as a sequential decision‐making problem.
Wenqian Yu, Jiaying Wu, Jinglu Hu
wiley   +1 more source