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/COSA

COSA

Reported on 21 benchmarks across 5 tasks · 1 paper

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

Computer Vision18 results

  • VideoonActivityNet
    text-to-video R@1· uses extra data· 2023-06-15
    67.3
    best: 74.1 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • VideoonDiDeMo
    text-to-video R@1· uses extra data· 2023-06-15
    70.5
    best: 74.2 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • VideoonMSR-VTT
    text-to-video R@1· uses extra data· 2023-06-15
    57.9
    best: 64 (GRAM)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • VideoonLSMDC
    text-to-video R@1· uses extra data· 2023-06-15
    39.4
    best: 46.4 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonMSR-VTT
    BLEU-4· uses extra data· 2023-06-15
    53.7
    best: 57.8 (mPLUG-2)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonMSR-VTT
    CIDEr· uses extra data· 2023-06-15
    74.7
    best: 80 (mPLUG-2)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonVATEX
    BLEU-4· uses extra data· 2023-06-15
    43.7
    best: 45.6 (VALOR)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonVATEX
    CIDEr· uses extra data· 2023-06-15
    96.5
    best: 99.5 (VAST)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonTVC
    BLEU-4· uses extra data· 2023-06-15
    18.8
    best: 19.9 (VAST)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonTVC
    CIDEr· uses extra data· 2023-06-15
    70.7
    best: 74.1 (VAST)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonYouCook2
    BLEU-4· uses extra data· 2023-06-15
    10.1
    best: 18.2 (VAST)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonYouCook2
    CIDEr· uses extra data· 2023-06-15
    1.31
    best: 116.4 (HowToCaption)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonMSVD
    BLEU-4· uses extra data· 2023-06-15
    76.5
    best: 80.7 (VALOR)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video CaptioningonMSVD
    CIDEr· uses extra data· 2023-06-15
    178.5
    best: 195.6 (MaMMUT)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video RetrievalonActivityNet
    text-to-video R@1· uses extra data· 2023-06-15
    67.3
    best: 74.1 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video RetrievalonDiDeMo
    text-to-video R@1· uses extra data· 2023-06-15
    70.5
    best: 74.2 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video RetrievalonMSR-VTT
    text-to-video R@1· uses extra data· 2023-06-15
    57.9
    best: 64 (GRAM)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video RetrievalonLSMDC
    text-to-video R@1· uses extra data· 2023-06-15
    39.4
    best: 46.4 (InternVideo2-6B)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085

Reasoning2 results

  • Video Question AnsweringonActivityNet-QA
    Accuracy· uses extra data· 2023-06-15
    49.9
    best: 61.6 (Tarsier (34B))
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085
  • Video Question AnsweringonMSRVTT-QA
    Accuracy· uses extra data· 2023-06-15
    49.2
    best: 72.4 (Flash-VStream)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085

Natural Language Processing1 result

  • Visual Question Answering (VQA)onMSVD-QA
    Accuracy· uses extra data· 2023-06-15
    0.6
    best: 0.61 (VLAB)
    COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelarXiv:2306.09085