Efficient Image Restoration via Latent Consistency Flow Matching
This work addresses the deployment challenge of image restoration on resource-constrained devices, representing an incremental improvement in efficiency over existing methods.
The paper tackles the problem of high computational demands in generative image restoration by introducing ELIR, an efficient latent method that balances distortion and perceptual quality, achieving a 4× reduction in size and speed compared to state-of-the-art approaches for tasks like blind face restoration.
Recent advances in generative image restoration (IR) have demonstrated impressive results. However, these methods are hindered by their substantial size and computational demands, rendering them unsuitable for deployment on edge devices. This work introduces ELIR, an Efficient Latent Image Restoration method. ELIR addresses the distortion-perception trade-off within the latent space and produces high-quality images using a latent consistency flow-based model. In addition, ELIR introduces an efficient and lightweight architecture. Consequently, ELIR is 4$\times$ smaller and faster than state-of-the-art diffusion and flow-based approaches for blind face restoration, enabling a deployment on resource-constrained devices. Comprehensive evaluations of various image restoration tasks and datasets show that ELIR achieves competitive performance compared to state-of-the-art methods, effectively balancing distortion and perceptual quality metrics while significantly reducing model size and computational cost. The code is available at: https://github.com/eladc-git/ELIR