Adaptive few-shot learning for robust part quality classification in two-photon lithography

2026-01-13

AI Summary

This work addresses the limitations of conventional computer vision models in two-photon lithography manufacturing, where dynamic environments, unknown defect detection, few-shot continual learning, and adaptation to novel part geometries remain challenging. To overcome these issues, the authors propose a full-lifecycle adaptive framework that integrates statistical hypothesis testing based on Linear Discriminant Analysis, a two-stage replay-based few-shot incremental learning strategy, and a few-shot Domain-Adversarial Neural Network (DANN), all built upon a scale-robust backbone architecture. The proposed method achieves 99–100% accuracy in detecting novel defect categories, integrates new defect classes with only 20 samples while maintaining 92% classification accuracy, and attains 96.19% cross-geometry transfer performance using merely five labeled samples in the target domain.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Two-photon lithography (TPL) is an advanced additive manufacturing (AM) technique for fabricating high-precision micro-structures. While computer vision (CV) is proofed for automated quality control, existing models are often static, rendering them ineffective in dynamic manufacturing environments. These models typically cannot detect new, unseen defect classes, be efficiently updated from scarce data, or adapt to new part geometries. To address this gap, this paper presents an adaptive CV framework for the entire life-cycle of quality model maintenance. The proposed framework is built upon a same, scale-robust backbone model and integrates three key methodologies: (1) a statistical hypothesis testing framework based on Linear Discriminant Analysis (LDA) for novelty detection, (2) a two-stage, rehearsal-based strategy for few-shot incremental learning, and (3) a few-shot Domain-Adversarial Neural Network (DANN) for few-shot domain adaptation. The framework was evaluated on a TPL dataset featuring hemisphere as source domain and cube as target domain structures, with each domain categorized into good, minor damaged, and damaged quality classes. The hypothesis testing method successfully identified new class batches with 99-100% accuracy. The incremental learning method integrated a new class to 92% accuracy using only K=20 samples. The domain adaptation model bridged the severe domain gap, achieving 96.19% accuracy on the target domain using only K=5 shots. These results demonstrate a robust and data-efficient solution for deploying and maintaining CV models in evolving production scenarios.

Problem

• few-shot learning
• quality classification
• two-photon lithography
• novelty detection
• domain adaptation

Innovation

• few-shot learning
• novelty detection
• incremental learning
• domain adaptation
• two-photon lithography