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/ViLT-B/32

ViLT-B/32

Reported on 39 benchmarks across 5 tasks · 1 paper · 18 SOTA

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

Miscellaneous36 results

Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@1· uses extra data· 2021-02-05
83.5
best: 98.8 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@10· uses extra data· 2021-02-05
98.6
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@5· uses extra data· 2021-02-05
96.7
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@10· uses extra data· 2021-02-05
93.8
best: 99.8 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@5· uses extra data· 2021-02-05
88.7
best: 99.5 (M2-Encoder)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@1· uses extra data· 2021-02-05
83.5
best: 98.8 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@10· uses extra data· 2021-02-05
98.6
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@5· uses extra data· 2021-02-05
96.7
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@1· uses extra data· 2021-02-05
64.4
best: 93.3 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@10· uses extra data· 2021-02-05
93.8
best: 99.8 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@5· uses extra data· 2021-02-05
88.7
best: 99.4 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@1· uses extra data· 2021-02-05
64.4
best: 93.3 (ERNIE-ViL 2.0)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@1· uses extra data· 2021-02-05
61.5
best: 84.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@10· uses extra data· 2021-02-05
92.7
best: 98.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@5· uses extra data· 2021-02-05
86.3
best: 96.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@1· uses extra data· 2021-02-05
42.7
best: 68 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@10· uses extra data· 2021-02-05
83.1
best: 92.8 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@5· uses extra data· 2021-02-05
72.9
best: 92.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@1· 2021-02-05
73.2
best: 98.8 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@10· 2021-02-05
96.5
best: 100 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@5· 2021-02-05
93.6
best: 100 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@1· 2021-02-05
55
best: 93.3 (ERNIE-ViL 2.0)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@10· 2021-02-05
89.8
best: 99.8 (ERNIE-ViL 2.0)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@5· 2021-02-05
82.5
best: 99.5 (M2-Encoder)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@1· 2021-02-05
56.5
best: 84.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@10· 2021-02-05
89.6
best: 98.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@5· 2021-02-05
82.6
best: 96.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@1· 2021-02-05
40.4
best: 68 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@10· 2021-02-05
81.1
best: 92.8 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@5· 2021-02-05
70
best: 92.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@1· uses extra data· 2021-02-05
61.5
best: 84.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@10· uses extra data· 2021-02-05
92.7
best: 98.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@5· uses extra data· 2021-02-05
86.3
best: 96.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@1· uses extra data· 2021-02-05
42.7
best: 68 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@10· uses extra data· 2021-02-05
83.1
best: 92.8 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@5· uses extra data· 2021-02-05
72.9
best: 92.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334

Natural Language Processing13 results

Cross-Modal RetrievalonFlickr30k
Image-to-text R@1· uses extra data· 2021-02-05
83.5
best: 98.8 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonFlickr30k
Image-to-text R@10· uses extra data· 2021-02-05
98.6
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonFlickr30k
Image-to-text R@5· uses extra data· 2021-02-05
96.7
best: 100 (X2-VLM (large))
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonFlickr30k
Text-to-image R@1· uses extra data· 2021-02-05
64.4
best: 93.3 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonFlickr30k
Text-to-image R@10· uses extra data· 2021-02-05
93.8
best: 99.8 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonFlickr30k
Text-to-image R@5· uses extra data· 2021-02-05
88.7
best: 99.4 (ERNIE-ViL 2.0)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Visual Question Answering (VQA)onVQA v2 test-dev
Accuracy· 2021-02-05
71.26
best: 84.3 (PaLI)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@1· uses extra data· 2021-02-05
61.5
best: 84.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@10· uses extra data· 2021-02-05
92.7
best: 98.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@5· uses extra data· 2021-02-05
86.3
best: 96.5 (X2-VLM (large))
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@1· uses extra data· 2021-02-05
42.7
best: 68 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@10· uses extra data· 2021-02-05
83.1
best: 92.8 (VAST)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@5· uses extra data· 2021-02-05
72.9
best: 92.8 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334

Reasoning2 results

Visual ReasoningonNLVR2 Dev
Accuracy· 2021-02-05
75.7
best: 91.51 (BEiT-3)
SOTA
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334
Visual ReasoningonNLVR2 Test
Accuracy· 2021-02-05
76.13
best: 92.58 (BEiT-3)
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision arXiv:2102.03334