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/MAE+MTP(ViT-B+RVSA)

MAE+MTP(ViT-B+RVSA)

Reported on 32 benchmarks across 10 tasks · 1 paper

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

Methodology19 results

Remote SensingonCDD Dataset (season-varying)
F1-Score· 2024-03-20
0.9787
best: 0.9842 (P2V-CD)
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Remote SensingonLEVIR-CD
F1· 2024-03-20
92.22
best: 92.67 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Remote SensingonLEVIR-CD
Params(M)· 2024-03-20
86
best: 335 (IMP+MTP(InternImage-XL))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
3DonDIOR
AP50· 2024-03-20
79.4
best: 81.1 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
3DonFAIR1M-2.0
mAP· 2024-03-20
51.92
best: 53 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
3DonxView
AP50· 2024-03-20
16.4
best: 19.4 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
3DonDIOR-R
mAP· 2024-03-20
71.29
best: 74.54 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D ClassificationonDIOR
AP50· 2024-03-20
79.4
best: 81.1 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D ClassificationonFAIR1M-2.0
mAP· 2024-03-20
51.92
best: 53 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D ClassificationonxView
AP50· 2024-03-20
16.4
best: 19.4 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D ClassificationonDIOR-R
mAP· 2024-03-20
71.29
best: 74.54 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D Object DetectiononDIOR
AP50· 2024-03-20
79.4
best: 81.1 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D Object DetectiononFAIR1M-2.0
mAP· 2024-03-20
51.92
best: 53 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D Object DetectiononxView
AP50· 2024-03-20
16.4
best: 19.4 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
2D Object DetectiononDIOR-R
mAP· 2024-03-20
71.29
best: 74.54 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
16konDIOR
AP50· 2024-03-20
79.4
best: 81.1 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
16konFAIR1M-2.0
mAP· 2024-03-20
51.92
best: 53 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
16konxView
AP50· 2024-03-20
16.4
best: 19.4 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
16konDIOR-R
mAP· 2024-03-20
71.29
best: 74.54 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430

Computer Vision9 results

Object DetectiononDIOR
AP50· 2024-03-20
79.4
best: 81.1 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Object DetectiononFAIR1M-2.0
mAP· 2024-03-20
51.92
best: 53 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Object DetectiononxView
AP50· 2024-03-20
16.4
best: 19.4 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Object DetectiononDIOR-R
mAP· 2024-03-20
71.29
best: 74.54 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Image ClassificationonEuroSAT
Accuracy (%)· 2024-03-20
98.76
best: 99.41 (DeepEnsembling)
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Change DetectiononWHU Building Dataset
F1-score· 2024-03-20
0.9432
best: 0.9559 (IMP+MTP(InternImage-XL))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Change DetectiononLEVIR-CD
F1· 2024-03-20
92.22
best: 92.87 (SChanger-base)
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Change DetectiononOSCD - 3ch
F1· 2024-03-20
53.36
best: 55.92 (MAE+MTP(ViT-L+RVSA))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Change DetectiononCDD Dataset (season-varying)
F1-Score· uses extra data· 2024-03-20
97.87
best: 98.42 (P2V-CD)
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430

Medical2 results

Semantic SegmentationonLoveDA
Category mIoU· 2024-03-20
52.39
best: 56.16 (U-Net (MaxViT-S))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
Semantic SegmentationonSpaceNet 1
Mean IoU· uses extra data· 2024-03-20
79.63
best: 79.69 (MAE+MTP(ViT-L))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430

Audio2 results

10-shot image generationonLoveDA
Category mIoU· 2024-03-20
52.39
best: 56.16 (U-Net (MaxViT-S))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430
10-shot image generationonSpaceNet 1
Mean IoU· uses extra data· 2024-03-20
79.63
best: 79.69 (MAE+MTP(ViT-L))
MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretraining arXiv:2403.13430