发表机构
CRRC Chongqing Co., Ltd.; College of Computer Science, Sichuan University; School of Artificial Intelligence, Hebei University of Technology; State Key Laboratory of Mechanical Transmission for Advanced Equipment, Chongqing University; College of Mathematics and Statistics, Chongqing University（中车重庆有限公司; 四川大学计算机学院; 河北工业大学人工智能学院; 重庆大学先进装备机械传动国家重点实验室; 重庆大学数学与统计学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对现有盲图像质量评估方法泛化能力受限的问题，提出基于拓扑不变流形学习的自监督BIQA框架，在铁路图像等场景实现了优异的零样本迁移与极端压力下的鲁棒性。
AI 中文摘要
现有的盲图像质量评估（BIQA）方法通常依赖合成失真和主观标注，限制了其在真实领域的泛化能力。为解决该问题，我们提出一种基于边界约束下拓扑不变流形学习的全自监督BIQA框架，无需人工标注即可构建稳定的质量参考。该框架通过围绕每个目标重复随机裁剪生成渐进式背景稀释尺度，利用这些尺度下目标信息密度的单调变化特性，建立自约束质量流形。线性空间矩投影消除随机裁剪带来的几何失真；随后，单调分歧滤波器修剪对背景敏感的评估器，分离出精英池\ud835\udefc_{\text{elite}}。带有主成分稳定器的稳健M估计器将这些指标融合为渐近高效的伪真值\ud870_{\text{PGT}}，使方差向克拉美罗下界收缩。大量评估表明，从11个基线指标中提炼的精英评估器池，在标准合成基准和野外基准（CSIQ、LIVEC、LIVE-2）上实现了卓越的零样本迁移能力；同时，在CQU铁路车辆监控数据集（2797张图像）上的部署结果显示，流形余弦相似度>0.999，在工业极端压力下的存活率达100.0%，有力验证了其跨范式解耦能力和拓扑鲁棒性。
英文摘要
Existing blind image quality assessment (BIQA) methods typically rely on synthetic distortions and subjective annotations, limiting generalization in real-world domains. To address this, we propose a fully self-supervised BIQA framework based on topologically invariant manifold learning under boundary constraints, which constructs a stable quality reference without manual labels. The framework generates progressive background dilution scales via repeated random cropping around each target; exploiting the monotonic degradation of target information density across these scales, it establishes a self-constrained quality manifold. A linearized spatial moment projection eliminates geometric distortions from random cropping; then a monotonicity divergence filter prunes background-sensitive evaluators, isolating an elite pool \(\mathcal{M}_{\text{elite}}\). A robust M-estimator with a principal component stabilizer fuses the metrics into an asymptotically efficient pseudo-ground truth \(q_{\text{PGT}}\), contracting variance toward the Cramér-Rao lower bound. Extensive evaluations demonstrate that the elite evaluator pool, distilled from 11 baseline metrics, secures superior zero-shot transferability across standard synthetic and wild benchmarks (CSIQ, LIVEC, LIVE-2). Concurrently, deployments on the CQU Railway Rolling Stock Surveillance Dataset (2,797 images) yield a manifold cosine similarity \(>0.999\) and a 100.0\% survival rate under industrial extreme stresses, robustly validating its cross-paradigm decoupling and topological resilience.
Comments13pages,14 tables, 5 figures
