Addressing Real-World Challenges in Recognition and Classification with Cutting-Edge AI Models and Methods

Dear Colleagues,

Deep learning has emerged as a transformative technology, significantly enhancing the performance of classification and recognition tasks across various domains. The ability of deep neural networks to extract complex patterns from large datasets has led to unprecedented breakthroughs in fields such as computer vision, natural language processing, language translations, speech recognition, unstructured language processing, medical diagnostics, drug discovery and intelligent automation.

Despite these advancements, challenges such as model interpretability, computational efficiency, data scarcity, and real-time deployment on edge devices continue to affect research. Additionally, with the growing complexity of classification problems, especially in biomedical AI, drug discovery and smart agriculture, there is a need to explore beyond classical deep learning approaches; this is primarily driven by limitations such as multilingual and unstructured data, and high-dimensional data representations.

Emerging paradigms such as quantum-enhanced machine learning (QML) and large language models (LLMs) offer new opportunities for the integration of adaptive knowledge, the optimization of feature representations, the acceleration of training, the enhancement of inference efficiency in data-scarce environments, and generalization across healthcare, agriculture, commonsense reasoning, biomedical applications, and programming.

This Special Issue aims to compile innovative research that presents the practical application of deep learning models in classification and recognition, addressing both fundamental challenges and novel approaches that enhance accuracy, efficiency, and real-world usability. More specifically, we welcome contributions whose scope includes, but is not limited to, the following challenges facing AI models and methods:

• Data Imbalance and Limited Annotations—Investigating novel techniques such as data augmentation, self-supervised learning, and active learning to address imbalanced and scarce datasets.
• Domain Adaptation and Transfer Learning—Exploring methods to improve model generalization across different datasets, domains, and real-world environments.
• Explainability and Interpretability in AI Models—Enhancing transparency in recognition and classification models for improved trust, fairness, and accountability.
• Few-Shot, Zero-Shot, and Self-Supervised Learning—Advancing recognition and classification models that require minimal labelled data for improved adaptability.
• Multimodal and Cross-Modal Learning—Integrating multiple data sources such as text, images, video, and sensor data to improve classification accuracy and robustness.
• Real-Time and Scalable AI Solutions—Optimizing recognition and classification models for high-speed, large-scale processing in resource-constrained environments.
• Bias Mitigation and Ethical AI—Addressing fairness, bias, and ethical concerns in AI-driven classification to ensure responsible and unbiased decision-making.
• Application-Specific Challenges and Innovations—Addressing domain-specific recognition problems in healthcare, agriculture, business, social media, and education using cutting-edge AI techniques.

Dr. Syed Naqvi
Dr. Md Mostafizer Rahman
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Information is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1800 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• classification and recognition
• large language models
• hybrid quantum–classical approaches
• multimodal and cross-modal learning
• class imbalance
• feature extraction and selection