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/SF-Net

SF-Net

Reported on 45 benchmarks across 5 tasks · 1 paper · 40 SOTA

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

Computer Vision36 results

  • VideoonGTEA
    mAP@0.1:0.7· 2020-03-15
    31
    best: 76.9 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonGTEA
    mAP@0.5· 2020-03-15
    19.3
    best: 66.3 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonBEOID
    mAP@0.1:0.7· 2020-03-15
    30.1
    best: 59.4 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonBEOID
    mAP@0.5· 2020-03-15
    16.7
    best: 55.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonTHUMOS 2014
    mAP@0.1:0.5· uses extra data· 2020-03-15
    51.2
    best: 71.6 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonTHUMOS 2014
    mAP@0.1:0.7· uses extra data· 2020-03-15
    41.2
    best: 60.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonTHUMOS 2014
    mAP@0.5· uses extra data· 2020-03-15
    30.5
    best: 52.2 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonActivityNet-1.2
    Mean mAP· 2020-03-15
    22.8
    best: 30.8 (SAL)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonGTEA
    mAP@0.1:0.7· 2020-03-15
    31
    best: 76.9 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonGTEA
    mAP@0.5· 2020-03-15
    19.3
    best: 66.3 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonBEOID
    mAP@0.1:0.7· 2020-03-15
    30.1
    best: 59.4 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonBEOID
    mAP@0.5· 2020-03-15
    16.7
    best: 55.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonTHUMOS 2014
    mAP@0.1:0.5· uses extra data· 2020-03-15
    51.2
    best: 71.6 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonTHUMOS 2014
    mAP@0.1:0.7· uses extra data· 2020-03-15
    41.2
    best: 60.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonTHUMOS 2014
    mAP@0.5· uses extra data· 2020-03-15
    30.5
    best: 52.2 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonActivityNet-1.2
    Mean mAP· 2020-03-15
    22.8
    best: 30.8 (SAL)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonGTEA
    mAP@0.1:0.7· 2020-03-15
    31
    best: 76.9 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonGTEA
    mAP@0.5· 2020-03-15
    19.3
    best: 66.3 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonBEOID
    mAP@0.1:0.7· 2020-03-15
    30.1
    best: 59.4 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonBEOID
    mAP@0.5· 2020-03-15
    16.7
    best: 55.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonTHUMOS 2014
    mAP@0.1:0.5· uses extra data· 2020-03-15
    51.2
    best: 71.6 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonTHUMOS 2014
    mAP@0.1:0.7· uses extra data· 2020-03-15
    41.2
    best: 60.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonTHUMOS 2014
    mAP@0.5· uses extra data· 2020-03-15
    30.5
    best: 52.2 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonActivityNet-1.2
    Mean mAP· 2020-03-15
    22.8
    best: 30.8 (SAL)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonGTEA
    mAP@0.1:0.7· 2020-03-15
    31
    best: 76.9 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonGTEA
    mAP@0.5· 2020-03-15
    19.3
    best: 66.3 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonBEOID
    mAP@0.1:0.7· 2020-03-15
    30.1
    best: 59.4 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonBEOID
    mAP@0.5· 2020-03-15
    16.7
    best: 55.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonTHUMOS 2014
    mAP@0.1:0.5· uses extra data· 2020-03-15
    51.2
    best: 71.6 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonTHUMOS 2014
    mAP@0.1:0.7· uses extra data· 2020-03-15
    41.2
    best: 60.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonTHUMOS 2014
    mAP@0.5· uses extra data· 2020-03-15
    30.5
    best: 52.2 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonActivityNet-1.2
    Mean mAP· 2020-03-15
    22.8
    best: 30.8 (SAL)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • VideoonActivityNet-1.2
    mAP@0.5· 2020-03-15
    37.8
    best: 49.6 (AICL)
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Temporal Action LocalizationonActivityNet-1.2
    mAP@0.5· 2020-03-15
    37.8
    best: 49.6 (AICL)
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Action LocalizationonActivityNet-1.2
    mAP@0.5· 2020-03-15
    37.8
    best: 49.6 (AICL)
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Weakly Supervised Action LocalizationonActivityNet-1.2
    mAP@0.5· 2020-03-15
    37.8
    best: 49.6 (AICL)
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845

Methodology9 results

  • Zero-Shot LearningonGTEA
    mAP@0.1:0.7· 2020-03-15
    31
    best: 76.9 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonGTEA
    mAP@0.5· 2020-03-15
    19.3
    best: 66.3 (AU-Action)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonBEOID
    mAP@0.1:0.7· 2020-03-15
    30.1
    best: 59.4 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonBEOID
    mAP@0.5· 2020-03-15
    16.7
    best: 55.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonTHUMOS 2014
    mAP@0.1:0.5· uses extra data· 2020-03-15
    51.2
    best: 71.6 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonTHUMOS 2014
    mAP@0.1:0.7· uses extra data· 2020-03-15
    41.2
    best: 60.3 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonTHUMOS 2014
    mAP@0.5· uses extra data· 2020-03-15
    30.5
    best: 52.2 (HR-Pro)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonActivityNet-1.2
    Mean mAP· 2020-03-15
    22.8
    best: 30.8 (SAL)
    SOTA
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845
  • Zero-Shot LearningonActivityNet-1.2
    mAP@0.5· 2020-03-15
    37.8
    best: 49.6 (AICL)
    SF-Net: Single-Frame Supervision for Temporal Action LocalizationarXiv:2003.06845