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/BEiT-3

BEiT-3

Reported on 52 benchmarks across 13 tasks · 1 paper · 35 SOTA

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

Miscellaneous30 results

Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@1· uses extra data· 2022-08-22
98
best: 98.8 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@1· uses extra data· 2022-08-22
90.3
best: 93.3 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@10· uses extra data· 2022-08-22
99.5
best: 99.8 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@5· uses extra data· 2022-08-22
98.7
best: 99.5 (M2-Encoder)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@1· uses extra data· 2022-08-22
84.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@10· uses extra data· 2022-08-22
98.3
best: 98.5 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Image-to-text R@5· uses extra data· 2022-08-22
96.5
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@1· uses extra data· 2022-08-22
67.2
best: 68 (VAST)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@5· uses extra data· 2022-08-22
92.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@1· uses extra data· 2022-08-22
98
best: 98.8 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@1· uses extra data· 2022-08-22
90.3
best: 93.3 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@10· uses extra data· 2022-08-22
99.5
best: 99.8 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Text-to-image R@5· uses extra data· 2022-08-22
98.7
best: 99.4 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@1· uses extra data· 2022-08-22
84.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@10· uses extra data· 2022-08-22
98.3
best: 98.5 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Image-to-text R@5· uses extra data· 2022-08-22
96.5
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@1· uses extra data· 2022-08-22
67.2
best: 68 (VAST)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@5· uses extra data· 2022-08-22
92.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@10· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@5· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonCOCO 2014
Text-to-image R@10· uses extra data· 2022-08-22
87.7
best: 92.8 (VAST)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@1· 2022-08-22
94.9
best: 98.8 (X2-VLM (large))
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@10· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Image-to-text R@5· 2022-08-22
99.9
best: 100 (X2-VLM (large))
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@1· 2022-08-22
81.5
best: 93.3 (ERNIE-ViL 2.0)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@10· 2022-08-22
97.8
best: 99.8 (ERNIE-ViL 2.0)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Image Retrieval with Multi-Modal QueryonFlickr30k
Text-to-image R@5· 2022-08-22
95.6
best: 99.5 (M2-Encoder)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@10· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonFlickr30k
Image-to-text R@5· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal Information RetrievalonCOCO 2014
Text-to-image R@10· uses extra data· 2022-08-22
87.7
best: 92.8 (VAST)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Natural Language Processing14 results

Visual Question Answering (VQA)onVQA v2 test-dev
Accuracy· 2022-08-22
84.19
best: 84.3 (PaLI)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Visual Question Answering (VQA)onVQA v2 test-std
overall· 2022-08-22
84.03
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Image-to-text R@1· uses extra data· 2022-08-22
98
best: 98.8 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Text-to-image R@1· uses extra data· 2022-08-22
90.3
best: 93.3 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Text-to-image R@10· uses extra data· 2022-08-22
99.5
best: 99.8 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Text-to-image R@5· uses extra data· 2022-08-22
98.7
best: 99.4 (ERNIE-ViL 2.0)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@1· uses extra data· 2022-08-22
84.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@10· uses extra data· 2022-08-22
98.3
best: 98.5 (X2-VLM (large))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Image-to-text R@5· uses extra data· 2022-08-22
96.5
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@1· uses extra data· 2022-08-22
67.2
best: 68 (VAST)
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@5· uses extra data· 2022-08-22
92.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Image-to-text R@10· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonFlickr30k
Image-to-text R@5· uses extra data· 2022-08-22
100
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Cross-Modal RetrievalonCOCO 2014
Text-to-image R@10· uses extra data· 2022-08-22
87.7
best: 92.8 (VAST)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Methodology4 results

3DonCOCO test-dev
box mAP· 2022-08-22
63.7
best: 66 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
2D ClassificationonCOCO test-dev
box mAP· 2022-08-22
63.7
best: 66 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
2D Object DetectiononCOCO test-dev
box mAP· 2022-08-22
63.7
best: 66 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
16konCOCO test-dev
box mAP· 2022-08-22
63.7
best: 66 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Medical3 results

Semantic SegmentationonADE20K val
mIoU· uses extra data· 2022-08-22
62.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Semantic SegmentationonADE20K
Validation mIoU· uses extra data· 2022-08-22
62.8
best: 63.6 (ViT-P (InternImage-H))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Semantic SegmentationonADE20K
Params (M)· uses extra data· 2022-08-22
1900
best: 3000 (FD-SwinV2-G)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Audio3 results

10-shot image generationonADE20K val
mIoU· uses extra data· 2022-08-22
62.8
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
10-shot image generationonADE20K
Validation mIoU· uses extra data· 2022-08-22
62.8
best: 63.6 (ViT-P (InternImage-H))
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
10-shot image generationonADE20K
Params (M)· uses extra data· 2022-08-22
1900
best: 3000 (FD-SwinV2-G)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Reasoning2 results

Visual ReasoningonNLVR2 Dev
Accuracy· 2022-08-22
91.51
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Visual ReasoningonNLVR2 Test
Accuracy· 2022-08-22
92.58
SOTA
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442

Computer Vision2 results

Object DetectiononCOCO test-dev
box mAP· 2022-08-22
63.7
best: 66 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442
Instance SegmentationonCOCO test-dev
mask AP· 2022-08-22
54.8
best: 57.1 (Co-DETR)
Image as a Foreign Language: BEiT Pretraining for All Vision and Vision-Language Tasks arXiv:2208.10442