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/SPAN

SPAN

Reported on 89 benchmarks across 16 tasks · 2 papers · 4 SOTA

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

Computer Vision41 results

  • Image Super-ResolutiononSet14 - 4x upscaling
    PSNR· 2023-11-21
    28.66
    best: 29.54 (DRCT-L)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • Image Super-ResolutiononSet14 - 4x upscaling
    SSIM· 2023-11-21
    0.7834
    best: 0.894 (Edge-informed SR)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • 3D Object Super-ResolutiononSet14 - 4x upscaling
    PSNR· 2023-11-21
    28.66
    best: 29.54 (DRCT-L)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • 3D Object Super-ResolutiononSet14 - 4x upscaling
    SSIM· 2023-11-21
    0.7834
    best: 0.894 (Edge-informed SR)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • Image Manipulation DetectiononCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Image Manipulation DetectiononCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Image Manipulation DetectiononCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Image Manipulation DetectiononCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Image Manipulation DetectiononDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Image Manipulation DetectiononDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Image Manipulation DetectiononCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Image Manipulation DetectiononCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)
  • VideoonCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • VideoonCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • VideoonCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • VideoonCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • VideoonDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • VideoonDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • VideoonCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • VideoonCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)
  • Temporal Action LocalizationonCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Temporal Action LocalizationonCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Temporal Action LocalizationonCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Temporal Action LocalizationonCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Temporal Action LocalizationonDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Temporal Action LocalizationonDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Temporal Action LocalizationonCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Temporal Action LocalizationonCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)
  • Action LocalizationonCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Action LocalizationonCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Action LocalizationonCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Action LocalizationonCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Action LocalizationonDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Action LocalizationonDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Action LocalizationonCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Action LocalizationonCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)
  • Image Manipulation LocalizationonColumbia
    Average Pixel F1(Fixed threshold)
    0.759
    best: 0.888 (Early Fusion)
  • Image Manipulation LocalizationonCOVERAGE
    Average Pixel F1(Fixed threshold)
    0.235
    best: 0.663 (Early Fusion)
  • Image Manipulation LocalizationonCasia V1+
    Average Pixel F1(Fixed threshold)
    0.112
    best: 0.791 (CMX (RGB+SRM))
  • Image Manipulation LocalizationonCocoGlide
    Average Pixel F1(Fixed threshold)
    0.298
    best: 0.585 (CMX (RGB+SRM))
  • Image Manipulation LocalizationonDSO-1
    Average Pixel F1(Fixed threshold)
    0.233
    best: 0.93 (TruFor)

Methodology18 results

  • 16konSet14 - 4x upscaling
    PSNR· 2023-11-21
    28.66
    best: 29.54 (DRCT-L)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • 16konSet14 - 4x upscaling
    SSIM· 2023-11-21
    0.7834
    best: 0.894 (Edge-informed SR)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • Anomaly DetectiononCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Anomaly DetectiononCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Anomaly DetectiononCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Anomaly DetectiononCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Anomaly DetectiononDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Anomaly DetectiononDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Anomaly DetectiononCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Anomaly DetectiononCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)
  • Zero-Shot LearningonCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Zero-Shot LearningonCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Zero-Shot LearningonCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Zero-Shot LearningonCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Zero-Shot LearningonDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Zero-Shot LearningonDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Zero-Shot LearningonCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Zero-Shot LearningonCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)

Natural Language Processing13 results

  • Sentiment AnalysisonSemEval 2014 Task 4 Subtask 1+2
    F1· 2019-06-10
    68.06
    best: 83.76 (gpt-3.5 finetuned)
    SOTA
    Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and ClassificationarXiv:1906.03820
  • Sentiment AnalysisonSemEval 2014 Task 4 Laptop
    F1· 2019-06-10
    68.06
    best: 79.34 (InstructABSA)
    SOTA
    Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and ClassificationarXiv:1906.03820
  • Aspect-Based Sentiment Analysis (ABSA)onSemEval 2014 Task 4 Subtask 1+2
    F1· 2019-06-10
    68.06
    best: 83.76 (gpt-3.5 finetuned)
    SOTA
    Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and ClassificationarXiv:1906.03820
  • Aspect-Based Sentiment Analysis (ABSA)onSemEval 2014 Task 4 Laptop
    F1· 2019-06-10
    68.06
    best: 79.34 (InstructABSA)
    SOTA
    Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and ClassificationarXiv:1906.03820
  • Sentiment AnalysisonSemEval 2014 Task 4 Subtask 1+2
    F1· 2019-06-10
    68.06
    best: 83.76 (gpt-3.5 finetuned)
    Open-Domain Targeted Sentiment Analysis via Span-Based Extraction and ClassificationarXiv:1906.03820
  • 3D Action RecognitiononCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • 3D Action RecognitiononCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • 3D Action RecognitiononCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • 3D Action RecognitiononCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • 3D Action RecognitiononDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • 3D Action RecognitiononDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • 3D Action RecognitiononCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • 3D Action RecognitiononCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)

Robots8 results

  • Activity RecognitiononCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Activity RecognitiononCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Activity RecognitiononCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Activity RecognitiononCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Activity RecognitiononDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Activity RecognitiononDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Activity RecognitiononCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Activity RecognitiononCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)

Time Series8 results

  • Action RecognitiononCOVERAGE
    AUC
    0.67
    best: 0.839 (Early Fusion)
  • Action RecognitiononCOVERAGE
    Balanced Accuracy
    0.235
    best: 0.77 (Early Fusion)
  • Action RecognitiononCocoGlide
    AUC
    0.475
    best: 0.778 (ManTraNet)
  • Action RecognitiononCocoGlide
    Balanced Accuracy
    0.298
    best: 0.677 (Late Fusion)
  • Action RecognitiononDSO-1
    AUC
    0.669
    best: 0.984 (TruFor)
  • Action RecognitiononDSO-1
    Balanced Accuracy
    0.233
    best: 0.935 (Early Fusion)
  • Action RecognitiononCasia V1+
    AUC
    0.48
    best: 0.942 (CAT-Net v2)
  • Action RecognitiononCasia V1+
    Balanced Accuracy
    0.112
    best: 0.86 (Late Fusion)

Graphs2 results

  • Super-ResolutiononSet14 - 4x upscaling
    PSNR· 2023-11-21
    28.66
    best: 29.54 (DRCT-L)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770
  • Super-ResolutiononSet14 - 4x upscaling
    SSIM· 2023-11-21
    0.7834
    best: 0.894 (Edge-informed SR)
    Swift Parameter-free Attention Network for Efficient Super-ResolutionarXiv:2311.12770