2026-10-08T04:33:39Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34032627 2026-09-30T14:51:34Z category_26650 category_29173 item_type_3 month_year_09_2026

Data for: Accelerating Two-Photon Lithography Prototyping through Rapid Cross-Modal Inspection Aditya Ghosh (25144191) Harnjoo Kim (16631886) Sourabh Saha (23833746) Additive manufacturing Machine learning not elsewhere classified Machine learning two-photon polymerization nanoprinting nanoscale 3D-printing Manufacturing inspection <p dir="ltr">This dataset contains the data generated for the paper:Accelerating Two-Photon Lithography Prototyping through Rapid Cross-Modal Inspection. </p><p dir="ltr">Specifically, it includes scanning electron microscopy images, optical microscope images, digital bitmap projection images, class labels, and processing conditions that were used to train deep learning machine learning models to automate the inspection of processing outcomes of grayscale projection two-photon lithography. </p><p dir="ltr">A summary of the work is: "Two-photon lithography (TPL) process optimization is limited by a metrology bottleneck: slow, destructive SEM-based verification. We present a rapid, resource-efficient inspection framework that uses only in-situ optical imaging while leveraging cross-modal machine learning to transfer metrological knowledge from SEM. Applied to grayscale projection-TPL, the framework employs dual-scale inspection: a structure-level model that screens complete samples with 95% grouped accuracy and 100% within-one-score accuracy on an eight-level ordinal scale, and an area-level model that detects localized polymerization defects with 88% accuracy. Transfer learning improves performance on unseen 3D woodpiles, reducing error by 79%, enabling scalable closed-loop nanoscale 3D printing"</p> 2026-09-30T14:51:34Z Dataset Dataset 10.6084/m9.figshare.34032627.v1 https://figshare.com/articles/dataset/Data_for_Accelerating_Two-Photon_Lithography_Prototyping_through_Rapid_Cross-Modal_Inspection/34032627 CC BY 4.0