New CoDe-SSM model enhances UHD image restoration efficiency

Researchers have introduced CoDe-SSM, a novel State Space Model designed for efficient Ultra-High-Definition (UHD) image restoration. This model separates the processing of aggregated context and localized details into distinct pathways. The context pathway uses a Global Cluster Scan Module (GCSM) to group features into cluster centers and apply selective SSM reasoning, allowing for cross-region context sharing independent of spatial resolution. The detail pathway, a Local High-Frequency Module (LHFM), reconstructs fine image structures using a high-frequency mask and sparse convolutional experts. Experiments on multiple UHD benchmarks show that this context-detail decoupling strategy significantly improves restoration quality while maintaining efficiency. AI

IMPACT Introduces a novel approach to image restoration that balances detail preservation with computational efficiency.

RANK_REASON Research paper detailing a new model for image restoration. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• CoDe-SSM
• Global Cluster Scan Module
• Hugging Face
• Local High-Frequency Module

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →