Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

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

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

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

Models/VinVL (Microsoft Cognitive Services + MSR)

VinVL (Microsoft Cognitive Services + MSR)

Reported on 32 benchmarks across 1 task · 1 paper · 8 SOTA

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

Natural Language Processing32 results

Image Captioningonnocaps entire
B1· 2021-01-02
81.59
best: 88.1 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
B2· 2021-01-02
65.15
best: 74.81 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
B3· 2021-01-02
45.04
best: 57.68 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
B4· 2021-01-02
26.15
best: 37.71 (CoCa - Google Brain)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
CIDEr· 2021-01-02
92.46
best: 126.8 (Lyrics)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
METEOR· 2021-01-02
27.57
best: 32.5 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
ROUGE-L· 2021-01-02
56.96
best: 63.12 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps entire
SPICE· 2021-01-02
13.07
best: 15.94 (GIT, Single Model)
SOTA
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
B1· 2021-01-02
82.77
best: 88.9 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
B2· 2021-01-02
66.94
best: 75.86 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
B3· 2021-01-02
47.02
best: 58.99 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
B4· 2021-01-02
27.97
best: 39.98 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
CIDEr· 2021-01-02
95.16
best: 125.51 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
METEOR· 2021-01-02
28.24
best: 33.47 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
ROUGE-L· 2021-01-02
57.95
best: 63.99 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps near-domain
SPICE· 2021-01-02
13.36
best: 16.11 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
B1· 2021-01-02
75.78
best: 86.28 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
B2· 2021-01-02
56.1
best: 71.28 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
B3· 2021-01-02
34.02
best: 52.66 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
B4· 2021-01-02
15.86
best: 32 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
CIDEr· 2021-01-02
78.01
best: 126.67 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
METEOR· 2021-01-02
23.55
best: 30.99 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
ROUGE-L· 2021-01-02
51.99
best: 61.35 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps out-of-domain
SPICE· 2021-01-02
11.48
best: 15.7 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
B1· 2021-01-02
83.24
best: 88.86 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
B2· 2021-01-02
68.04
best: 76.1 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
B3· 2021-01-02
49.68
best: 60.53 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
B4· 2021-01-02
30.62
best: 41.65 (GIT, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
CIDEr· 2021-01-02
97.99
best: 149.1 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
METEOR· 2021-01-02
29.51
best: 34.22 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
ROUGE-L· 2021-01-02
58.54
best: 64.39 (PaLI)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529
Image Captioningonnocaps in-domain
SPICE· 2021-01-02
13.63
best: 16.36 (GIT2, Single Model)
VinVL: Revisiting Visual Representations in Vision-Language Models arXiv:2101.00529