arXiv:2605.29212v1 Announce Type: new 
Abstract: Image quality in modern imaging systems emerges from the coupled effects of the sensor, optics, and computational reconstruction. Ultra-thin metalenses offer a path toward substantial miniaturization of optical modules, but practical designs often exhibit pronounced chromatic and field-dependent aberrations that necessitate computational reconstruction. In current metalens pipelines, reconstruction models are commonly trained and selected using distortion-based fidelity objectives, such as PSNR, yet these proxies can be weakly correlated with human preference and downstream utility, reflecting the well-known perception--distortion trade-off. We introduce MetaRanker, a human-in-the-loop active ranking framework that formalizes metalens image quality in terms of semantic interpretability, defined as the degree to which humans can reliably recognize objects and structures in the presence of optical artifacts. MetaRanker combines a probabilistic preference model with uncertainty-aware query selection, and leverages vision--language models to provide lightweight semantic priors. Importantly, these priors are used only to guide the sampling of informative comparisons; human judgments remain the primary supervision signal throughout. Across real-world and synthetic metalens datasets with distinct degradation profiles, MetaRanker produces rankings that align most closely with human assessments, while reducing the number of pairwise annotations required by approximately 80% relative to exhaustive pairwise evaluation. Finally, we show that standard image quality assessment metrics exhibit limited alignment with human interpretability in the metalens domain, positioning MetaRanker as a practical step toward perceptually grounded metalens evaluation and co-design.

تم تقديم MetaRanker كإطار تصنيف نشط بمشاركة بشرية يهدف إلى تحسين جودة الصورة في العدسات المعدنية، وهي أجهزة بصرية فائقة النحافة. يتناول هذا الإطار قيود الأهداف القائمة على التشويه من خلال التركيز على القابلية للتفسير الدلالي، مما يسمح بالتعرف الأفضل على الكائنات على الرغم من العيوب البصرية.

MetaRanker se ha presentado como un marco de clasificación activa con intervención humana, destinado a mejorar la calidad de imagen en metalentes, que son dispositivos ópticos ultradelgados. Este marco aborda las limitaciones de los objetivos de fidelidad basados en la distorsión al centrarse en la interpretabilidad semántica, permitiendo un mejor reconocimiento de objetos a pesar de los artefactos ópticos.

MetaRanker a été introduit comme un cadre de classement actif impliquant l'humain, visant à améliorer la qualité d'image des metalenses, qui sont des dispositifs optiques ultra-fins. Ce cadre aborde les limitations des objectifs de fidélité basés sur la distorsion en se concentrant sur l'interprétabilité sémantique, permettant une meilleure reconnaissance des objets malgré les artefacts optiques.

MetaRanker has been introduced as a human-in-the-loop active ranking framework aimed at improving image quality in metalenses, which are ultra-thin optical devices. This framework addresses the limitations of traditional distortion-based fidelity objectives by focusing on semantic interpretability, allowing for better recognition of objects despite optical artifacts.