Towards Real-world Lens Active Alignment with Unlabeled Data via Domain Adaptation

2026-01-07

arXiv.org

AI Summary

This work addresses the domain shift between simulation and real-world scenarios in automated optical system assembly by proposing the DA3 framework. DA3 leverages a self-supervised learning strategy that combines an autoregressive domain translation generator with adversarial feature alignment to extract domain-invariant degradation features from a small set of unlabeled real images. This approach effectively bridges the gap between simulated and real imaging conditions while substantially reducing reliance on labeled real-world data. Experimental validation on two lens types demonstrates that, compared to purely simulation-based training, DA3 improves active alignment accuracy by 46%, achieving performance comparable to fully supervised methods while cutting real data acquisition time by 98.7%.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Active Alignment (AA) is a key technology for the large-scale automated assembly of high-precision optical systems. Compared with labor-intensive per-model on-device calibration, a digital-twin pipeline built on optical simulation offers a substantial advantage in generating large-scale labeled data. However, complex imaging conditions induce a domain gap between simulation and real-world images, limiting the generalization of simulation-trained models. To address this, we propose augmenting a simulation baseline with minimal unlabeled real-world images captured at random misalignment positions, mitigating the gap from a domain adaptation perspective. We introduce Domain Adaptive Active Alignment (DA3), which utilizes an autoregressive domain transformation generator and an adversarial-based feature alignment strategy to distill real-world domain information via self-supervised learning. This enables the extraction of domain-invariant image degradation features to facilitate robust misalignment prediction. Experiments on two lens types reveal that DA3 improves accuracy by 46% over a purely simulation pipeline. Notably, it approaches the performance achieved with precisely labeled real-world data collected on 3 lens samples, while reducing on-device data collection time by 98.7%. The results demonstrate that domain adaptation effectively endows simulation-trained models with robust real-world performance, validating the digital-twin pipeline as a practical solution to significantly enhance the efficiency of large-scale optical assembly.

Problem

• Active Alignment
• Domain Adaptation
• Unlabeled Data
• Optical Assembly
• Simulation-to-Real Gap

Innovation

• Domain Adaptation
• Active Alignment
• Digital Twin
• Self-supervised Learning
• Optical Assembly