About Me
Matin Fazel is a Master’s student in Computer Science at Concordia University, based in Montréal, Canada. He is currently a Research Assistant at the In2GM Lab, where he works under the supervision of Abdelhak Bentaleb. His work sits at the intersection of multimedia systems, networking, and applied machine learning, with a strong focus on real-time and browser-based technologies.
During his graduate studies, he has led two research papers at the ACM MMSys (rank-A) conference, while contributing to ongoing research efforts within the lab. His research emphasizes efficient real-time video processing, WebRTC-based systems, and GPU-accelerated inference, particularly in the context of real-time streaming and adaptive multimedia applications.
As part of his work, Matin has designed and developed systems such as a browser-native video conferencing platform enhanced with real-time super-resolution, as well as telemetry-driven frameworks for dynamic system optimization. His projects demonstrate a strong combination of systems engineering and machine learning, including large-scale performance instrumentation and low-latency GPU pipelines.
Beyond research, Matin has industry experience as a software developer, where he built scalable backend services and real-time multimedia systems, including an enterprise monitoring dashboard and a high-performance video streaming server.
Prior to his master’s, he completed his Bachelor’s in Computer Engineering at Isfahan University of Technology, where he also worked on machine learning applications such as speech emotion augmentation using generative models.
Selected Publications
- VSR-Bench: An Open-Source Platform for Browser-Native Real-Time VSR Evaluation in WebRTC, ACM MMSys 2026. Introduces the first browser-native benchmarking platform for real-time video super-resolution in WebRTC, enabling controlled evaluation of latency, perceptual quality, and GPU execution trade-offs. Code
- More Pixels, Less Bandwidth: A Live Demo of VSR-Bench over WebRTC, ACM MMSys 2026. Presents a live interactive demo of receiver-side video super-resolution in WebRTC, showing how model choice, runtime configuration, and network conditions affect end-to-end quality of experience.
News
- 🏆 April 7, 2026: Received the Best Demo Award at the 18th ACM Multimedia Systems Conference (MMSys 2026) for the paper “More Pixels, Less Bandwidth: A Live Demo of VSR-Bench over WebRTC”
- 🥇 April 3, 2026: Model RFDN_SPAN qualified and ranked 7th at The Eleventh NTIRE 2026 Efficient Super-Resolution Challenge at CVPR 2026 NTIRE Workshop
- 📄 February 16, 2025: Paper VSR-Bench: An Open-Source Platform for Browser-Native Real-Time VSR Evaluation in WebRTC accepted at ACM MMSys ‘26
- 📄 February 14, 2025: Paper More Pixels, Less Bandwidth: A Live Demo of VSR-Bench over WebRTC accepted at ACM MMSys ‘26
- 🎓 September 02, 2024: Started Master’s study at Concordia University
- ✨ February 16, 2024: Got accepted at Concordia University in the Master of Computer Science and received the Concordia Merit Scholarship
- 🎉 February 04, 2024: Finished Bachelor’s study from Isfahan University of Technology
