Joint Geometric and Trajectory Consistency Learning for One-Step Real-World Super-Resolution

2026-02-27

AI Summary

This work addresses the structural distortions prevalent in existing single-step real-world super-resolution methods based on consistency models, which suffer from accumulated consistency drift and geometric decoupling. To tackle this, we propose the GTASR framework, which is the first to explicitly identify and mitigate the “geometric decoupling” problem. Our approach introduces trajectory alignment (TA) via full-path projection to correct the tangent vector field and incorporates a dual-reference structure refinement (DRSR) mechanism to effectively suppress drift and enhance geometric fidelity. Extensive experiments demonstrate that GTASR achieves state-of-the-art visual quality across multiple real-world benchmarks while maintaining extremely low inference latency.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Diffusion-based Real-World Image Super-Resolution (Real-ISR) achieves impressive perceptual quality but suffers from high computational costs due to iterative sampling. While recent distillation approaches leveraging large-scale Text-to-Image (T2I) priors have enabled one-step generation, they are typically hindered by prohibitive parameter counts and the inherent capability bounds imposed by teacher models. As a lightweight alternative, Consistency Models offer efficient inference but struggle with two critical limitations: the accumulation of consistency drift inherent to transitive training, and a phenomenon we term"Geometric Decoupling"- where the generative trajectory achieves pixel-wise alignment yet fails to preserve structural coherence. To address these challenges, we propose GTASR (Geometric Trajectory Alignment Super-Resolution), a simple yet effective consistency training paradigm for Real-ISR. Specifically, we introduce a Trajectory Alignment (TA) strategy to rectify the tangent vector field via full-path projection, and a Dual-Reference Structural Rectification (DRSR) mechanism to enforce strict structural constraints. Extensive experiments verify that GTASR delivers superior performance over representative baselines while maintaining minimal latency. The code and model will be released at https://github.com/Blazedengcy/GTASR.

Problem

• Consistency Models
• Real-World Super-Resolution
• Geometric Decoupling
• Consistency Drift
• Structural Coherence

Innovation

• Consistency Models
• Geometric Decoupling
• Trajectory Alignment
• Real-World Super-Resolution
• One-Step Generation