aiqbench analysis
This research presents a novel metalens architecture that combines spectral filtering with neural network optimization to overcome image degradation caused by obstructions like rain or dust. By selectively focusing light from distant objects while filtering near-depth interference, the system achieves significant gains in object detection accuracy and semantic segmentation without relying on bulky mechanical cleaning. The approach advances compact vision systems for drones and robots, demonstrating a tangible convergence of advanced optics and machine learning to solve practical sensing challenges.
De-occluding broadband metalens
arXiv:2601.19403v2 Announce Type: replace-cross Abstract: Obstructions such as raindrops, fences, or dust degrade captured images, especially when mechanical cleaning is infeasible. Conventional solutions to obstructions rely on a bulky compound optics array or computational inpainting, which compromise compactness or fidelity. Metalenses composed of subwavelength meta-atoms promise compact imaging, but simultaneous achievement of broadband and obstruction-free imaging remains a challenge, since a metalens that images distant scenes across a broadband spectrum cannot properly defocus near-depth occlusions. Here, we introduce a learned split-spectrum metalens that enables broadband obstruction-free imaging. Our approach divides the spectrum of each RGB channel into pass and stop bands with multi-band spectral filtering and learns the metalens to focus light from far objects through pass bands, while filtering focused near-depth light through stop bands. This optical signal is further enhanced using a neural network. Our learned split-spectrum metalens achieves broadband and obstruction-free imaging with relative PSNR gains of 32.29% and improves object detection and semantic segmentation accuracies with absolute gains of +13.54% mAP, +48.45% IoU, and +20.35% mIoU over a conventional hyperbolic design. This promises robust obstruction-free sensing and vision for space-constrained systems, such as mobile robots, drones, and endoscopes.
Abstract:Obstructions such as raindrops, fences, or dust degrade captured images, especially when mechanical cleaning is infeasible. Conventional solutions to obstructions rely on a bulky compound optics array or computational inpainting, which compromise compactness or fidelity. Metalenses composed of subwavelength meta-atoms promise compact imaging, but simultaneous achievement of broadband and obstruction-free imaging remains a challenge, since a metalens that images distant scenes across a broadband spectrum cannot properly defocus near-depth occlusions. Here, we introduce a learned split-spectrum metalens that enables broadband obstruction-free imaging. Our approach divides the spectrum of each RGB channel into pass and stop bands with multi-band spectral filtering and learns the metalens to focus light from far objects through pass bands, while filtering focused near-depth light through stop bands. This optical signal is further enhanced using a neural network. Our learned split-spectrum metalens achieves broadband and obstruction-free imaging with relative PSNR gains of 32.29% and improves object detection and semantic segmentation accuracies with absolute gains of +13.54% mAP, +48.45% IoU, and +20.35% mIoU over a conventional hyperbolic design. This promises robust obstruction-free sensing and vision for space-constrained systems, such as mobile robots, drones, and endoscopes.
Optics (physics.optics); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Applied Physics (physics.app-ph)
arXiv:2601.19403 [physics.optics]
(or arXiv:2601.19403v2 [physics.optics] for this version)
https://doi.org/10.48550/arXiv.2601.19403 Focus to learn more arXiv-issued DOI via DataCite
https://doi.org/10.1038/s41467-026-76606-0 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 300 words
Related coverage
Evidence of Absence: Cross-Modal Abductive Risk Perception to Sustain World Models When Vision Fails
arXiv · Aug 18, 2026
Mask-Induced Displacement in Audio XAI via Logit Trajectory Decomposition
arXiv · Sep 30, 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · Sep 24, 2026
Metalens scans entire eye in 3D with 1.2mm depth
Quantum Zeitgeist · Sep 4, 2026