Real Time Video Upscaling with AI: A Hardware-Aware Deep Learning Approaches

Author(s)	Mr. Aryan Kankheria, Prof. Dr. Srisakthi Saravanan
Country	India
Abstract	Efficient real-time image upscaling is essential for enhancing visual quality in gaming and real-world applications, especially on hardware-limited devices. In this work, I present a customized version of ESRGAN, trained on diverse real-world and gaming datasets, optimized for real-time performance and high-quality upscaling. The trained model is converted into an ONNX format to ensure broad compatibility while maintaining low latency. Our method achieves a balance between inference speed and visual fidelity, outperforming other industry-standard algorithms in both perceptual quality and objective metrics. Experimental results demonstrate significant improvements in resolution enhancement, texture sharpness, and artifact reduction, making the proposed approach suitable for deployment on resource-constrained hardware.
Keywords	Computer Vision, AI, Super Resolution, Real-Time Upscaling, Deep learning, ESRGAN, FSRCNN, LAPSRN, CUDA, ONNX Model, Custom Dataset, Unreal Engine 4
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 7, Issue 5, September-October 2025
Published On	2025-10-21
DOI	https://doi.org/10.36948/ijfmr.2025.v07i05.49690

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research