Results 201 to 210 of about 5,574 (249)
Visualizing Image Segmentation Network Behavior Through the Lens of Scale Space Analysis
Abstract Deep neural networks are widely used for image segmentation, also in sensitive applications such as medical imaging or autonomous driving. However, few explainable AI methods are available that help developers understand such networks beyond classification.
A. C. Mikliss, T. Schultz
wiley +1 more source
SPINE: VAE‐driven Counterfactuals for Decision Boundary Maps
Abstract As Deep Learning models become increasingly complex, Explainable AI becomes essential for deploying machine learning classifiers. Decision Boundary Mapping (DBM) is a technique for visualizing a classifier's global decision boundary. Despite their relative success, current DBM methods rely on global inverse multidimensional projections that ...
I.M. Bloemen, V. Prasad, F. V. Paulovich
wiley +1 more source
Threats and vulnerabilities in artificial intelligence and agentic AI models. [PDF]
Radanliev P, Santos O, Maple C.
europepmc +1 more source
FaceIt: Real‐Time On‐Device Identity‐Preserving Portrait Animation
A real‐time, on‐device algorithm for animating a single portrait image using detailed facial expressions and motions from a driving video achieving visual quality that can outperform even techniques with significantly higher computational overhead. Abstract Recently, significant progress has been made in generating expressive and temporally coherent ...
M. Dvorožňák +7 more
wiley +1 more source
The specifics of non‐routine task changes: A granular approach
Abstract This paper estimates granular specific non‐routine tasks and examines employment reallocation across those tasks in the United States. Employment shifts into decision‐making, technology, and information‐related tasks account for 90% of high‐wage employment growth.
Carol A. Scotese
wiley +1 more source
Hybrid GAN-LSTM framework for diabetic foot ulcer image synthesis and automated diagnosis. [PDF]
Vina A +7 more
europepmc +1 more source
Textile and colour defect detection using deep learning methods
Abstract Recent advances in deep learning (DL) have significantly enhanced the detection of textile and colour defects. This review focuses specifically on the application of DL‐based methods for defect detection in textile and coloration processes, with an emphasis on object detection and related computer vision (CV) tasks.
Hao Cui +2 more
wiley +1 more source
Toward leveraging intrinsic point cloud features in 3D adversarial attacks. [PDF]
Naderi H, Dinesh C, Bajić IV, Kasaei S.
europepmc +1 more source
Abstract This article explores the complexity of interorganizational relationships in social entrepreneurship. Although existing research has deepened our understanding of collaboration, there remains a need to examine the broader range of relationships social entrepreneurs navigate in practice. To address this issue, we adopt an ecosystem perspective,
Wiebke Heinze, Jarrod Ormiston
wiley +1 more source
Query-Efficient Hard-Label Attack: A Prior-Guided Adam Ray Search Optimization. [PDF]
Ding T, Xu X, Xuan Q, Yu H, Ma C.
europepmc +1 more source

