Aims & Scope

The International Journal of Pattern Recognition and Artificial Intelligence (IJPRAI) welcomes both theory-oriented and innovative applications articles on new developments and is of interest to both researchers in academia and industry.

The current scope of this journal includes:

• Pattern Recognition
• Machine Learning
• Deep Learning
• Document Analysis
• Image Processing
• Signal Processing
• Computer Vision
• Biometrics
• Biomedical Image Analysis
• Artificial Intelligence

In addition to regular papers describing original research work, survey articles on timely and important research topics are highly welcome. Special issues with focused topics within the scope of this journal are also published.

Special Issue on Lightweight and Energy-Efficient Artificial Intelligence: Innovation in Edge Intelligence for Pattern Recognition and Generative Models
提出日: 2026-05-30

Edge Intelligence, a key direction of IoT-AI integration, drives fields like intelligent transportation, smart security, and industrial IoT toward "terminal autonomous decision-making." Yet edge devices face constraints in computing resources, energy efficiency, and real-time performance. Traditional deep learning models—with large parameters and high computational costs—are hard to deploy efficiently at the edge. Against this, lightweight and energy-efficient AI has become critical to overcoming edge intelligence implementation bottlenecks. As core AI pillars, pattern recognition and generative models still face integration challenges at the edge: pattern recognition (e.g., real-time edge detection) needs low-power accuracy, while mature lightweight design and efficient inference for generative models (e.g., image generation) remain lacking, making their collaborative optimization and cross-task integration urgent issues.

Exploring their innovative edge applications has three key values: Technically, it advances integration of lightweight AI (model compression, knowledge distillation, dynamic computing) with edge tasks, breaking resource-constrained performance limits. Application-wise, it empowers edge devices with integrated "perception-decision-generation" intelligence, accelerating edge evolution from "single-point perception" to "global collaboration" and offering cost-effective solutions for IoT terminals and wearables. Disciplinarily, it fosters cross-field innovation among pattern recognition, generative models, and edge computing, building an "algorithm lightweighting—task efficiency—scenario generalization" paradigm and filling lightweight AI gaps in edge multi-task fusion.

This special issue focuses on "lightweight and energy-efficient AI," delving into edge intelligence innovation for pattern recognition and generative models. It offers a platform for academia and industry, advancing theoretical breakthroughs and applications of lightweight AI in edge scenarios to boost edge intelligence innovation and ecosystem development. Original research on theory, algorithms, experiments, and applications is welcome, focusing on these topics:

Dynamic Computing Scheduling of Generative Networks and Performance Balance in Pattern Recognition
Adaptive Inference of Generative Models and Pattern Recognition Applications in Edge Environments
Pattern Recognition Innovation of Lightweight Generative Models in Edge-Based Industrial Defect Detection
Iterative Learning and Interpretability Enhancement of Neuro-Symbolic AI in Lightweight Generative Models
Pattern Recognition and Semantic Understanding via Neuro-Symbolic Fusion for Multimodal Edge Data
Collaborative Optimization Strategies for Energy Efficiency-Accuracy in Pattern Recognition Under Edge Computing Environments
Resource Scheduling and Task Adaptation of Lightweight AI Models in Edge IoT Perception
Hardware-Algorithm Co-Design for Real-Time Pattern Recognition Algorithms on Low-Power Edge Devices
Meta-Computing Driven Improvement of Generalization and Consistency for Cross-Edge-Device Pattern Recognition Models
Adaptation of Meta-Learning Based Lightweight Pattern Recognition Models in Dynamic Edge Scenarios
Privacy Preservation and Performance Trade-Off of Pattern Recognition and Generative Models Under Edge Federated Learning
Collaboration of Lightweight Models for Edge Medical Image Generation and Recognition and Exploration of Clinical Applications

Please note that submissions received from 15 January 2026 onwards will be published in Open Access upon acceptance, with an Article Processing Charge (APC) of US$2,200 per article.

Submission Deadlines:

Manuscript Submission Deadline: 30.05.2026
Authors Notification: 15.07.2026
Revised Papers Due: 30.08.2026
Final Notification: 15.10.2026