{
  "abstract": "Few-shot adaptation of vision–language models (VLMs) could reduce the labeling effort for defect screening in scanning electron microscopy (SEM) of polymer electrolyte membrane (PEM) fuel-cell materials. We built a leakage-controlled benchmark from the PEMDefect-1107 corpus to measure what VLM language priors contribute. Its controls (size-matched crops, grouping of augmented copies, validation-only selection, common calibration, 20 support draws, image-level bootstrap) were needed: with naive crops, five low-level statistics separated the classes with an area under the receiver operating characteristic curve (AUROC) of 0.951. Zero-shot prompting reached at most 0.685 Macro-F1. With 16 labeled patches per class, text-initialized adaptation of CLIP ViT-L/14 reached 0.823 and, relative to a linear probe on the same features, reduced the labels needed to reach 95% of a fine-tuned ResNet-50 from 512 to 32 per class. A linear probe on ImageNet ResNet-50 features did not differ significantly from it at any tested label budget (difference at 16 shots −0.005; 95% confidence interval −0.059 to 0.044), and two microscopy-pretrained encoders were weaker. On whole micrographs, the better 16-shot classifiers placed 55–63% of the defects in the top 10% of the area. Within each VLM, the text prototypes raised the mean AUROC but gave no detectable advantage over a strong generic encoder.",
  "abstract_shown": true,
  "affiliations": [
    "Department of Energy Systems Engineering, Fırat University, Elazığ 23119, Türkiye",
    "Department of Electrical and Electronics Engineering, Kahramanmaraş Sütçü İmam University, Kahramanmaraş 46050, Türkiye"
  ],
  "article_type": null,
  "arxiv": null,
  "author_total": 3,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "Sami Ekici",
      "orcid": "0000-0002-6760-2183"
    },
    {
      "affs": [
        1
      ],
      "name": "Turab Selçuk",
      "orcid": "0000-0002-8445-2105"
    },
    {
      "affs": [
        0
      ],
      "name": "Masud Kabir",
      "orcid": "0000-0001-9766-226X"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/ed1bdy4vc826b.bib",
  "cited_by": 0,
  "doi": "10.3390/batteries12110450",
  "fulltext_url": null,
  "id": "ed1bdy4vc826b",
  "issns": [
    "2313-0105"
  ],
  "landing_url": "https://www.mdpi.com/2313-0105/12/11/450",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [],
  "openalex": null,
  "pdf_url": null,
  "pmcid": null,
  "pmid": null,
  "published": "2026-11-02",
  "publisher": "MDPI AG",
  "ris": "https://research.draconitic.com/paper/ed1bdy4vc826b.ris",
  "sources": [
    {
      "first_seen": "2026-11-04T00:24:21Z",
      "native_id": "10.3390/batteries12110450",
      "source": "crossref",
      "url": "https://www.mdpi.com/2313-0105/12/11/450"
    }
  ],
  "subjects": [
    {
      "field": "chemistry",
      "origin": "issn:2313-0105",
      "subfield": "electrochemistry-energy"
    }
  ],
  "title": "Benchmarking Vision–Language and Vision Foundation Models for Few-Shot Defect Screening in PEM Fuel-Cell SEM Micrographs",
  "title_text": "Benchmarking Vision–Language and Vision Foundation Models for Few-Shot Defect Screening in PEM Fuel-Cell SEM Micrographs",
  "type": "peer-reviewed",
  "updated_at": "2026-11-04T00:24:21Z",
  "url": "https://research.draconitic.com/paper/ed1bdy4vc826b",
  "venue": "Batteries",
  "versions": []
}