Proposed a test-time scaling paradigm for image restoration that allows compute-to-quality trade-offs during inference.
March 24, 2026
Original Paper
Tuning Real-World Image Restoration at Inference: A Test-Time Scaling Paradigm for Flow Matching Models
arXiv · 2603.22027
The Takeaway
It brings the 'scaling at inference' trend from LLMs to computer vision, using a reward model to dynamically steer Flow Matching models. This enables significant performance gains on image restoration without the need for expensive model retraining.
From the abstract
Although diffusion-based real-world image restoration (Real-IR) has achieved remarkable progress, efficiently leveraging ultra-large-scale pre-trained text-to-image (T2I) models and fully exploiting their potential remain significant challenges. To address this issue, we propose ResFlow-Tuner, an image restoration framework based on the state-of-the-art flow matching model, FLUX.1-dev, which integrates unified multi-modal fusion (UMMF) with test-time scaling (TTS) to achieve unprecedented restor