NanoVSR: Towards Real-Time Video Super-Resolution on Edge Devices

📅 2026-07-11
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
This work addresses the high computational cost of existing video super-resolution methods, which typically rely on Transformers or explicit optical flow estimation, hindering deployment on edge devices. To overcome this limitation, the authors propose NanoVSR, a lightweight fully convolutional architecture that leverages structural reparameterization to employ only standard convolutions during inference. Combined with a progressive training strategy, NanoVSR implicitly learns spatiotemporal alignment without explicit motion compensation, achieving high-quality super-resolution while maintaining efficiency and hardware acceleration compatibility. Experimental results demonstrate that NanoVSR-644k attains 28.64 dB PSNR on REDS4 and achieves real-time performance (27.2 FPS) on a Jetson Orin NX; the larger NanoVSR-1.7M variant further improves PSNR to 29.15 dB while sustaining 19.58 FPS.
📝 Abstract
Recent Video Super-Resolution (VSR) methods rely heavily on transformers and explicit optical flow, creating computational overhead and custom operations that hinder deployment on hardware accelerators like TensorRT. To address this, we introduce NanoVSR, a scalable, fully convolutional architecture designed for resource-constrained edge devices. Using structural reparameterization, NanoVSR collapses into standard convolutions during inference, ensuring seamless hardware compatibility and negligible runtime overhead. Furthermore, despite lacking explicit motion compensation, it maintains competitive restoration quality by implicitly learning spatio-temporal alignments through progressive training. Evaluated on the REDS4 benchmark, NanoVSR demonstrates an exceptional balance between accuracy and computational efficiency, significantly improving the trade-off for compact architectures. Our NanoVSR-644k baseline yields 28.64 dB PSNR while delivering 27.2 FPS on the NVIDIA Jetson Orin NX 16GB (25W), offering massive speed gains over heavier models. The scaled NanoVSR-1.7M variant reaches 29.15 dB with a throughput of 19.58 FPS, providing superior, edge-optimized upscaling. Code is available at https://github.com/filippawlicki/nanovsr.
Problem

Research questions and friction points this paper is trying to address.

Video Super-Resolution
Edge Devices
Real-Time
Hardware Deployment
Computational Overhead
Innovation

Methods, ideas, or system contributions that make the work stand out.

structural reparameterization
fully convolutional architecture
implicit spatio-temporal alignment
edge-optimized video super-resolution
real-time inference
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
F
Filip Pawlicki
Gdańsk University of Technology, Faculty of Electronics, Telecommunications and Informatics, Department of Computer Architecture, Gdańsk, Poland
M
Marcel Kańduła
Gdańsk University of Technology, Faculty of Electronics, Telecommunications and Informatics, Department of Software Engineering, Gdańsk, Poland
M
Marcin Pucek
Gdańsk University of Technology, Faculty of Electronics, Telecommunications and Informatics, Department of Computer Architecture, Gdańsk, Poland
K
Kamil Dobies
Gdańsk University of Technology, Faculty of Electronics, Telecommunications and Informatics, Department of Computer Architecture, Gdańsk, Poland