TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/GLASS

GLASS

Reported on 17 benchmarks across 2 tasks · 2 papers · 9 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology12 results

  • Anomaly DetectiononMPDD
    Detection AUROC· 2024-07-12
    99.6
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononMPDD
    Segmentation AUPRO· 2024-07-12
    98.2
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononMPDD
    Segmentation AUROC· 2024-07-12
    99.4
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononWFDD
    Detection AUROC· 2024-07-12
    100
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononWFDD
    Segmentation AUPRO· 2024-07-12
    94.9
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononWFDD
    Segmentation AUROC· 2024-07-12
    98.9
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononMVTec AD
    Detection AUROC· uses extra data· 2024-07-12
    99.9
    SOTA
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononMVTec AD
    Segmentation AUPRO· uses extra data· 2024-07-12
    96.8
    best: 98.4 (WeakREST-Block)
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononMVTec AD
    Segmentation AUROC· uses extra data· 2024-07-12
    99.3
    best: 99.7 (WeakREST-Block)
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononVisA
    Detection AUROC· 2024-07-12
    98.8
    best: 99.8 (UniNet)
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononVisA
    Segmentation AUPRO (until 30% FPR)· 2024-07-12
    92.8
    best: 96.1 (AnomalyDINO-S (full-shot))
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359
  • Anomaly DetectiononVisA
    Segmentation AUROC· 2024-07-12
    98.8
    best: 99.1 (Dinomaly ViT-L (model-unified multi-class))
    A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and LocalizationarXiv:2407.09359

Computer Vision5 results

  • Text SpottingonTotal-Text
    F-measure (%) - No Lexicon· 2022-08-05
    76.6
    best: 83.6 (DeepSolo (ViTAEv2-S, TextOCR))
    SOTA
    GLASS: Global to Local Attention for Scene-Text SpottingarXiv:2208.03364
  • Text SpottingonICDAR 2015
    F-measure (%) - Generic Lexicon· 2022-08-05
    76.3
    best: 80.3 (UNITS)
    SOTA
    GLASS: Global to Local Attention for Scene-Text SpottingarXiv:2208.03364
  • Text SpottingonTotal-Text
    F-measure (%) - Full Lexicon· 2022-08-05
    83
    best: 89.6 (DeepSolo (ViTAEv2-S, TextOCR))
    GLASS: Global to Local Attention for Scene-Text SpottingarXiv:2208.03364
  • Text SpottingonICDAR 2015
    F-measure (%) - Strong Lexicon· 2022-08-05
    84.7
    best: 89 (UNITS)
    GLASS: Global to Local Attention for Scene-Text SpottingarXiv:2208.03364
  • Text SpottingonICDAR 2015
    F-measure (%) - Weak Lexicon· 2022-08-05
    80.1
    best: 84.1 (UNITS)
    GLASS: Global to Local Attention for Scene-Text SpottingarXiv:2208.03364