Submission Portal Open: October 27, 2024
Submission Deadline: March 31, 2025

========================

=== Call for papers ===

The rapid advancement of visual pattern recognition systems has led to
their transition from laboratory settings to real-world applications,
where they face the challenges of distribution shifts and adversarial
samples. This special issue focuses on innovative methodologies that
enhance the robustness and generalization capabilities of visual
classifiers on unknown data in diverse, uncontrolled environments,
addressing key issues such as dataset imbalance, adversarial attacks,
and the exploitation of multi-modal systems. Submissions are encouraged
from researchers exploring neural network architectures, data
augmentation, multi-task learning, and multi-sensor fusion techniques to
improve performance in real-world conditions.

This special issue seeks to collect cutting-edge research that advances
the generalization capabilities of visual classifiers under real-world
conditions. The scope includes, but is not limited to, the development
of robust neural network architectures, transformers, and machine
learning models that address challenges such as distribution shift,
adversarial attacks, and dataset imbalance. Contributions leveraging
multi-task neural networks, multimodal approaches (e.g., vision-language
models, multi-sensor fusion), and efficient, lightweight models for edge
devices are highly encouraged. Papers should align with the broader
topics of computer vision, image processing, multimedia systems, and
biometrics, with a focus on improving real-world performance across
various applications, including autonomous driving, cognitive robotics,
and security-critical environments.

Topics of interest are but not limited to:

1) Novel Neural Networks or other Architectures (e.g. Transformers) for
Dealing with Distribution Shifts in the Wild
2) Data Augmentation Strategies, Generative and Degradation models for
Enhancing Generalization on Unseen Data
3) Robustness against Adversarial Attacks
4) Bias Mitigation in Unbalanced Datasets
5) Multi-task vs Single-task Learning in Real-world Scenarios
6) Resource-efficient Architectures for Edge Computing and (near)
Real-time Processing
7) Vision-Language Models and other Multi-modal Approaches
8) Multi-sensor Fusion for Enhanced Performance
9) New Datasets and Benchmarks for Computer Vision Systems in the Wild
10) Novel Applications and Case Studies

========================

=== Guest editors ===

George Azzopardi, PhD
University of Groningen, Groningen, The Netherlands
E-mail: g.azzopardi@rug.nl

Laura Fernández Robles, PhD
University of León, Leon, Spain
E-mail: l.fernandez@unileon.es

Antonio Greco, PhD
University of Salerno, Fisciano, Italy
E-mail: agreco@unisa.it

Bruno Vento, PhD Student
University of Naples Federico II, Napoli, Italy
E-mail: bruno.vento@unina.it