This two-volume set CCIS 2858-2859 constitutes the refereed proceedings of the 5th International Conference on Intelligent Systems and Pattern Recognition, ISPR 2025, held in Hammamet, Tunisia, during September 25–27, 2025.

The 59 full papers presented in these volumes were carefully reviewed and selected from the 180 submissions. They over a wide spectrum of contemporary research topics, including machine learning, deep learning, computer vision, data mining, intelligent systems, explainable artificial intelligence, pattern recognition methodologies, and multimedia analysis. 

.- Spine 3-D Representation from 2-D X-Ray Images and Measurement for Scoliosis Patient 
Follow-Up.
.- Stacked Deep Learning Models Leveraging Denoising Autoencoder-Based Features for 
Student Performance Prediction.
.- A Deep Attention Bi-LSTM Pipeline for -Human Action Recognition(HAR).
.- Balancing Accuracy and Efficiency in Deep Violence Detection Models for Surveillance 
Applications.
.- Comparative Analysis of Deep Learning and Classical Machine Learning for Deepfake 
Image Detection.
.- Automatic Discretization of Depression Symptoms from Social Media Using KMeans and 
Quantiles.
.- ProdNet: A Lightweight Network for Fast Discovery of Matrix Multiplication Algorithms.
.- Hybrid CNN-Transformer Model for Ovarian Cancer Lesion Classification.
.- Comparative Analysis of SAC and PPO for Energy Management in Battery Electric Vehicles.
.- On-Device Lightweight CNNs for Industrial Thermal Fault Diagnosis and Monitoring.
.- Multi-Views Knee MRI Classification Using Vision Transformers for Automated Meniscal 
Tear Detection.
.- A Case Study on Applying StyleGAN2-ADA for Fashion Image Synthesis with Limited Data.
.- A U-Net GNN Hybrid Approach for MRI-Based Brain Tumor Segmentation.
.- HeteroGAT: A Heterogeneous Graph Attention Network for Multimodal Fake News 
Detection in Chinese Social Media.
.- An Event-Based Data Generation Framework for Autonomous Driving Scenarios.
.- Improving Low-Resource Dialect Recognition Through Audio Data Augmentation: A Case 
Study on Algerian Language.
.- From Farm to Warehouse: Toward a Semantic and AI-Based Architecture for Wheat 
Lifecycle.
.- Spatio-Temporal Transformers for High-Accuracy Detection of Embryo Developmental 
Transitions.
.- Prompting Recovery: Filling Missing Diabetes Data with Large Language Models.
.- Regularized Continual Learning for Generative Crowd Counting.
.- Real-Time Defect Detection: A Lightweight Deep Learning Framework for Industrial 
Applications.
.- Robust 1D CNN-Based Approach with Data Augmentation for Incipient and Severe Stator 
Fault Classification in Induction Motors.
.- A Comparative Study of Arabic Embedding Models in RAG-Based Fatwa Retrieval.
.- Sentiment Classification of COVID-19 Tweets: From Machine Learning and Deep Learning 
to BERT.
.- Maintenance-Oriented Test Generators and Test Teams Simulation for Demand-Controlled 
Ventilation and Heating Systems.
.- CASCADE: Engineering a Secure, Real-Time, and Collaborative Code Execution for Multi
Tenant Learning Platforms.
.- A Review of E-Document Analysis with Multi-Biometric Technique for Secure Verification.
.- NeuroYOLO: Data-Efficient Fine-Tuning of YOLOv11 for Vision-Based Rehabilitation 
Monitoring.
.- Machine Learning-based Autism Detection using Oriented Basic Image Features.