CoChat
Sign Up
Cited by 585Industrial Vision Systems and Defect DetectionAnomaly Detection Techniques and ApplicationsDomain Adaptation and Few-Shot Learning

The DOI matches its Crossref record; 56 of the 58 references it cites were found in a registry or an index.

WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation

Open access
Jeong, Jongheon, Zou, Yang, Kim, Taewan, Zhang, Dongqing, Ravichandran, Avinash, Dabeer, Onkar·2023·2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
DOI 10.1109/cvpr52729.2023.01878
Open this paper in CoChat — free accountOpen PDF

Abstract

Visual anomaly classification and segmentation are vital for automating industrial quality inspection. The focus of prior research in the field has been on training custom models for each quality inspection task, which requires task-specific images and annotation. In this paper we move away from this regime, addressing zero-shot and few-normal-shot anomaly classification and segmentation. Recently CLIP, a vision-language model, has shown revolutionary generality with competitive zero-/few-shot performance in comparison to full-supervision. But CLIP falls short on anomaly classification and segmentation tasks. Hence, we propose window-based CLIP (WinCLIP) with (1) a compositional ensemble on state words and prompt templates and (2) efficient extraction and aggregation of window/patch/image-level features aligned with text. We also propose its few-normal-shot extension Win-CLIP+, which uses complementary information from normal images. In MVTec-AD (and VisA), without further tuning, WinCLIP achieves 91.8%/85.1% (78.1%/79.6%) AU-ROC in zero-shot anomaly classification and segmentation while WinCLIP + does 93.1%/95.2% (83.8%/96.4%) in 1-normal-shot, surpassing state-of-the-art by large margins.

Overview

Integrity signals

Registry

Crossref · 10.1109/cvpr52729.2023.01878

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Citations

Cites

Show all 58 references

Research works printed but not linked: 4

  1. geoffrey, 2012, ImageNet classification with deep convolutional neural networks, Advances in neural information processing systems, 25
  2. ilharco, 2021, Open cl
  3. 10.1109/CVPR46437.2021.01466 · DOI · Paper page
  4. alayrac, 2022, Flamingo: A visuallanguage model for few-shot learning, ArXiv Preprint

Cited by

Sorted by citations

Items 501–585 of 585

Related

See all

Metadata

Authors
Jeong, J., Zou, Y., Kim, T., Zhang, D., Ravichandran, A., & Dabeer, O.
Affiliations
  • Korea Advanced Institute of Science and Technology (KR)
  • Amazon (United States) (US)
ORCID
Venue
2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Publisher
IEEE
Published
2023-06
Pages
19606-19616
ISSN
1063-6919 (ISSN-L)
DOI
10.1109/cvpr52729.2023.01878
OpenAlex
W4386065385
Semantic Scholar
aa207668318fec38d60b79f407fb64982e46fce9
Cited by
585
References
58
Citations by year
2026: 286 · 2025: 206 · 2024: 89 · 2023: 4
FWCI
253.73
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Engineering › Industrial and Manufacturing Engineering › Industrial Vision Systems and Defect Detection
  • Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
  • Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
Keywords
anomaly segmentation, zero-shot learning, few-shot learning, CLIP, vision-language models, prompt templates, industrial inspection, MVTec AD, Visa

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review