Learning Single-Image Super-Resolution in the JPEG Compressed Domain
By: Sruthi Srinivasan , Elham Shakibapour , Rajy Rawther and more
Potential Business Impact:
Makes AI learn faster by skipping image decoding.
Deep learning models have grown increasingly complex, with input data sizes scaling accordingly. Despite substantial advances in specialized deep learning hardware, data loading continues to be a major bottleneck that limits training and inference speed. To address this challenge, we propose training models directly on encoded JPEG features, reducing the computational overhead associated with full JPEG decoding and significantly improving data loading efficiency. While prior works have focused on recognition tasks, we investigate the effectiveness of this approach for the restoration task of single-image super-resolution (SISR). We present a lightweight super-resolution pipeline that operates on JPEG discrete cosine transform (DCT) coefficients in the frequency domain. Our pipeline achieves a 2.6x speedup in data loading and a 2.5x speedup in training, while preserving visual quality comparable to standard SISR approaches.
Similar Papers
Dataset Distillation for Super-Resolution without Class Labels and Pre-trained Models
CV and Pattern Recognition
Makes blurry pictures sharp with less data.
Two-Stage Vision Transformer for Image Restoration: Colorization Pretraining + Residual Upsampling
CV and Pattern Recognition
Makes blurry pictures sharp and clear.
Dual-domain Adaptation Networks for Realistic Image Super-resolution
CV and Pattern Recognition
Makes blurry photos sharp and clear.