Work

Enhancing personalized video streaming through contextual information and support for emerging video formats

Year: 2025

Type: article

Abstract: This study presents an intelligent personalized video streaming framework that integrates contextual awareness and support for emerging video formats such as HDR and $$360^{\circ }$$ videos. Using a deep learning-based QoE prediction model, the proposed system adapts bitrate and resolution dynamically based on network conditions, device capabilities, and user preferences. Experimental results across diverse video formats show that our approach achieves up to 17.4% improvement in QoE compared to baseline models (NARX, Decision Tree, and Linear Regression), with statistically significant differences (p < 0.05). The framework supports deployment across edge and cloud infrastructure with consideration for real-time constraints. These results demonstrate the feasibility of our system for scalable, real-world adaptive streaming.

Source: Cluster Computing

DOI: 10.1007/s10586-025-05621-5

Authors: Mahmoud Darwich, Magdy Bayoumi

Institutions: University of Mount Union, University of Louisiana at Lafayette

Language: English

Topic: Image and Video Quality Assessment

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: hybrid

Estimated APC paid: $3,190