Search papers, labs, and topics across Lattice.
This paper introduces the Uncertainty-guided Adverse-weather Restoration Network (UAR-Net), which utilizes a gated transformer architecture to effectively restore images degraded by various adverse weather conditions. By employing Gated Dual-scale Transformer Blocks and a Balanced Multi-scale Skip Connection, the model adeptly handles spatially heterogeneous degradations while incorporating an Uncertainty-Aware Refinement Head for enhanced detail and artifact removal. Experimental results indicate that UAR-Net outperforms existing methods on multiple benchmarks, showcasing its potential for robust image restoration in challenging environments.
UAR-Net achieves state-of-the-art image restoration from adverse weather by effectively modeling both global interactions and local structures, all while estimating predictive uncertainty.
Restoring images degraded by adverse weather remains challenging due to spatially heterogeneous degradations. Many existing weather-specific restoration models rely on weather-agnostic global aggregation, naive cross-scale fusion, and deterministic objectives, which struggle to handle heterogeneous degradations in all-in-one adverse-weather settings. To address these limitations, we propose an Uncertainty-guided Adverse-weather Restoration Network (UAR-Net), a weather-specific AiO framework that integrates a gated transformer with balanced multi-scale skip connections. Specifically, we employ Gated Dual-scale Transformer Blocks (GDTB) to jointly model selective global interactions and multi-scale local structures, a progressive Balanced Multi-scale Skip Connection (BMSC) for balanced multi-scale feature integration, and an Uncertainty-Aware Refinement Head (URH) that performs artifact removal, detail enhancement, and predictive uncertainty estimation. The model is supervised by a Brightness-Aware Energy Loss (BAE-Loss) to encourage accurate reconstruction with well-calibrated uncertainty. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple adverse-weather benchmarks. The codes will open source upon acceptance.
