Ubi-Media 2027 Special Session on Advanced AOI Techniques
(SS_AAOI 2027)

Organizers

Session Chair: Dr. Shih-Yu Chen

National Yunlin University of Science and Technology

Abstract

Recent advances in deep learning, hyperspectral & multispectral sensing, and edge computing have accelerated Automated Optical Inspection (AOI) well beyond factory floors. This Special Session welcomes work that pushes AOI from algorithmic innovation to robust, validated systems — across industrial quality control, food & agriculture, and clinical/medical use-cases. We emphasize papers that address real-world constraints: sensor calibration, lighting, limited labels (few-shot), real-time edge inference, multimodal fusion (RGB + spectral + depth), deployment case studies, and clinical validation pathways including ethics and regulatory considerations. Submissions describing prototypes, pilot clinical studies, production-line integrations, datasets, or reproducible benchmarks are especially encouraged. Our goal is to foster cross-disciplinary exchange between imaging engineers, AI researchers, clinicians, and industry practitioners to accelerate safe and measurable impact.

Scope & Topics of Interest includes —but not limited to:

  • Deep-learning architectures for defect / lesion detection (2D, 3D, temporal)

  • Hyperspectral / multispectral imaging applications (band selection, spectral–spatial fusion)

  • Snapshot vs line-scan tradeoffs for real-time imaging (industrial & clinical form-factors)

  • Few-shot, semi-/unsupervised, anomaly detection approaches for low-label regimes

  • Edge inference, quantization, hardware acceleration (embedded platforms, FPGA, TPU)

  • Multimodal fusion: RGB + spectral + depth + thermal + fluorescence + OCT

  • Imaging calibration, illumination design, cross-site standardization procedures

  • Clinical validation & trial design, IRB/ethics reporting, patient privacy & consent practices

  • Regulatory & quality-management considerations (FDA/CE pathways, risk analysis)

  • Case studies: production-line integration, intraoperative guidance, endoscopy, dermatology, digital pathology

  • Dataset creation, annotation protocols, benchmark tasks, reproducibility resources

Financial Co-Chairs

  • Yung-Hui Chen, LungHwa University of Sci. and Tech., Taiwan

Submission & Format information:

Please submit your papers via EasyChair at:
https://easychair.org/conferences/?conf=ss-aaoi2027)