Efficient Perceptual Image Super Resolution: AIM 2025 Study and Benchmark
This work addresses the problem of inefficient perceptual image super-resolution for practical deployment, representing an incremental improvement with a new benchmark.
The paper tackled the inefficiency of perceptual super-resolution methods by developing solutions that outperform Real-ESRGAN across all datasets while adhering to strict constraints of 5M parameters and 2000 GFLOPs, using a novel 500-image benchmark.
This paper presents a comprehensive study and benchmark on Efficient Perceptual Super-Resolution (EPSR). While significant progress has been made in efficient PSNR-oriented super resolution, approaches focusing on perceptual quality metrics remain relatively inefficient. Motivated by this gap, we aim to replicate or improve the perceptual results of Real-ESRGAN while meeting strict efficiency constraints: a maximum of 5M parameters and 2000 GFLOPs, calculated for an input size of 960x540 pixels. The proposed solutions were evaluated on a novel dataset consisting of 500 test images of 4K resolution, each degraded using multiple degradation types, without providing the original high-quality counterparts. This design aims to reflect realistic deployment conditions and serves as a diverse and challenging benchmark. The top-performing approach manages to outperform Real-ESRGAN across all benchmark datasets, demonstrating the potential of efficient methods in the perceptual domain. This paper establishes the modern baselines for efficient perceptual super resolution.