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/UNINEXT-H

UNINEXT-H

Reported on 79 benchmarks across 13 tasks · 1 paper · 30 SOTA

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

Computer Vision55 results

  • VideoonBDD100K val
    mIDF1· 2023-03-12
    56.7
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual TrackingonTNL2K
    AUC· 2023-03-12
    59.3
    best: 60.3 (ARTrack-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual TrackingonTNL2K
    precision· 2023-03-12
    62.8
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT-ext
    AUC· 2023-03-12
    56.2
    best: 61 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT-ext
    Normalized Precision· 2023-03-12
    63.8
    best: 73.9 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT-ext
    Precision· 2023-03-12
    63.8
    best: 72.2 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonTrackingNet
    Precision· 2023-03-12
    86.4
    best: 89.2 (MCITrack-L384)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonBDD100K val
    mIDF1· 2023-03-12
    56.7
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefCoCo val
    Overall IoU· uses extra data· 2023-03-12
    82.19
    best: 85.41 (DeRIS-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefer-YouTube-VOS (2021 public validation)
    F· 2023-03-12
    72.7
    best: 76.1 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefer-YouTube-VOS (2021 public validation)
    J· 2023-03-12
    67.6
    best: 71.7 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefer-YouTube-VOS (2021 public validation)
    J&F· 2023-03-12
    70.1
    best: 73.9 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefCOCO+ val
    Overall IoU· uses extra data· 2023-03-12
    72.47
    best: 79.4 (MLCD-Seg-7B)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefCOCO+ test B
    Overall IoU· uses extra data· 2023-03-12
    66.22
    best: 75.6 (MLCD-Seg-7B)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonDAVIS 2017 (val)
    J&F 1st frame· uses extra data· 2023-03-12
    72.5
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonRefCOCO+ testA
    Overall IoU· uses extra data· 2023-03-12
    76.42
    best: 83.5 (HyperSeg)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefCoCo val
    Overall IoU· uses extra data· 2023-03-12
    82.19
    best: 85.41 (DeRIS-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefer-YouTube-VOS (2021 public validation)
    F· 2023-03-12
    72.7
    best: 76.1 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefer-YouTube-VOS (2021 public validation)
    J· 2023-03-12
    67.6
    best: 71.7 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefer-YouTube-VOS (2021 public validation)
    J&F· 2023-03-12
    70.1
    best: 73.9 (MPG-SAM 2)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefCOCO+ val
    Overall IoU· uses extra data· 2023-03-12
    72.47
    best: 79.4 (MLCD-Seg-7B)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefCOCO+ test B
    Overall IoU· uses extra data· 2023-03-12
    66.22
    best: 75.6 (MLCD-Seg-7B)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonDAVIS 2017 (val)
    J&F 1st frame· uses extra data· 2023-03-12
    72.5
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Referring Expression SegmentationonRefCOCO+ testA
    Overall IoU· uses extra data· 2023-03-12
    76.42
    best: 83.5 (HyperSeg)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Multi-Object Tracking and SegmentationonBDD100K val
    mMOTSA· 2023-03-12
    35.7
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Multiple Object TrackingonBDD100K val
    mIDF1· 2023-03-12
    56.7
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT-ext
    AUC· 2023-03-12
    56.2
    best: 61 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT-ext
    Normalized Precision· 2023-03-12
    63.8
    best: 73.9 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT-ext
    Precision· 2023-03-12
    63.8
    best: 72.2 (SAMURAI-L)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonTrackingNet
    Precision· 2023-03-12
    86.4
    best: 89.2 (MCITrack-L384)
    SOTA
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • VideoonBDD100K val
    mMOTA· 2023-03-12
    44.2
    best: 45.5 (ByteTrack)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT
    AUC· 2023-03-12
    72.2
    best: 77.4 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT
    Normalized Precision· 2023-03-12
    80.8
    best: 86.6 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonLaSOT
    Precision· 2023-03-12
    79.4
    best: 85 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonTrackingNet
    Accuracy· 2023-03-12
    85.4
    best: 87.9 (MCITrack-L384)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonTrackingNet
    Normalized Precision· 2023-03-12
    89
    best: 92.1 (MCITrack-L384)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object TrackingonBDD100K val
    mMOTA· 2023-03-12
    44.2
    best: 45.5 (ByteTrack)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    AP50· uses extra data· 2023-03-12
    77.5
    best: 82.1 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    AP75· uses extra data· 2023-03-12
    66.7
    best: 71.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    APL· uses extra data· 2023-03-12
    75.3
    best: 78.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    APM· uses extra data· 2023-03-12
    64.8
    best: 68.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    APS· uses extra data· 2023-03-12
    45.1
    best: 50.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Object DetectiononCOCO minival
    box AP· uses extra data· 2023-03-12
    60.6
    best: 66 (PE_spatial (DETA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    AP50· uses extra data· 2023-03-12
    76.2
    best: 80.8 (InternImage-H)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    AP75· uses extra data· 2023-03-12
    56.7
    best: 63.4 (Co-DETR)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    APL· uses extra data· 2023-03-12
    67.5
    best: 72.4 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    APM· uses extra data· 2023-03-12
    55.9
    best: 60.1 (Co-DETR)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    APS· uses extra data· 2023-03-12
    33.3
    best: 41.6 (Co-DETR)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Instance SegmentationonCOCO test-dev
    mask AP· uses extra data· 2023-03-12
    51.8
    best: 57.1 (Co-DETR)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Multiple Object TrackingonBDD100K val
    mMOTA· 2023-03-12
    44.2
    best: 45.5 (ByteTrack)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT
    AUC· 2023-03-12
    72.2
    best: 77.4 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT
    Normalized Precision· 2023-03-12
    80.8
    best: 86.6 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonLaSOT
    Precision· 2023-03-12
    79.4
    best: 85 (SPMTrack-G)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonTrackingNet
    Accuracy· 2023-03-12
    85.4
    best: 87.9 (MCITrack-L384)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • Visual Object TrackingonTrackingNet
    Normalized Precision· 2023-03-12
    89
    best: 92.1 (MCITrack-L384)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674

Methodology24 results

  • 3DonCOCO minival
    AP50· uses extra data· 2023-03-12
    77.5
    best: 82.1 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 3DonCOCO minival
    AP75· uses extra data· 2023-03-12
    66.7
    best: 71.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 3DonCOCO minival
    APL· uses extra data· 2023-03-12
    75.3
    best: 78.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 3DonCOCO minival
    APM· uses extra data· 2023-03-12
    64.8
    best: 68.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 3DonCOCO minival
    APS· uses extra data· 2023-03-12
    45.1
    best: 50.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 3DonCOCO minival
    box AP· uses extra data· 2023-03-12
    60.6
    best: 66 (PE_spatial (DETA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    AP50· uses extra data· 2023-03-12
    77.5
    best: 82.1 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    AP75· uses extra data· 2023-03-12
    66.7
    best: 71.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    APL· uses extra data· 2023-03-12
    75.3
    best: 78.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    APM· uses extra data· 2023-03-12
    64.8
    best: 68.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    APS· uses extra data· 2023-03-12
    45.1
    best: 50.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D ClassificationonCOCO minival
    box AP· uses extra data· 2023-03-12
    60.6
    best: 66 (PE_spatial (DETA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    AP50· uses extra data· 2023-03-12
    77.5
    best: 82.1 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    AP75· uses extra data· 2023-03-12
    66.7
    best: 71.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    APL· uses extra data· 2023-03-12
    75.3
    best: 78.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    APM· uses extra data· 2023-03-12
    64.8
    best: 68.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    APS· uses extra data· 2023-03-12
    45.1
    best: 50.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 2D Object DetectiononCOCO minival
    box AP· uses extra data· 2023-03-12
    60.6
    best: 66 (PE_spatial (DETA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    AP50· uses extra data· 2023-03-12
    77.5
    best: 82.1 (EVA)
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    AP75· uses extra data· 2023-03-12
    66.7
    best: 71.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    APL· uses extra data· 2023-03-12
    75.3
    best: 78.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    APM· uses extra data· 2023-03-12
    64.8
    best: 68.5 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    APS· uses extra data· 2023-03-12
    45.1
    best: 50.4 (Focal-Stable-DINO (Focal-Huge, no TTA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674
  • 16konCOCO minival
    box AP· uses extra data· 2023-03-12
    60.6
    best: 66 (PE_spatial (DETA))
    Universal Instance Perception as Object Discovery and RetrievalarXiv:2303.06674