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/SwinMTL

SwinMTL

Reported on 18 benchmarks across 6 tasks · 1 paper · 14 SOTA

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

Methodology11 results

Transfer LearningonNYUv2
Mean IoU· 2024-03-15
58.14
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Transfer LearningonCityscapes test
RMSE· 2024-03-15
0.51
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Transfer LearningonCityscapes test
mIoU· 2024-03-15
76.41
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
3DonCityscapes test
RMSE· 2024-03-15
6.352
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
3DonCityscapes
Absolute relative error (AbsRel)· 2024-03-15
0.089
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
3DonCityscapes
RMSE· 2024-03-15
5.481
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
3DonCityscapes
RMSE log· 2024-03-15
0.139
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Multi-Task LearningonNYUv2
Mean IoU· 2024-03-15
58.14
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Multi-Task LearningonCityscapes test
RMSE· 2024-03-15
0.51
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Multi-Task LearningonCityscapes test
mIoU· 2024-03-15
76.41
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
3DonCityscapes
Square relative error (SqRel)· 2024-03-15
1.051
best: 0.792 (Manydepth2)
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662

Computer Vision5 results

Depth EstimationonCityscapes test
RMSE· 2024-03-15
6.352
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Depth EstimationonCityscapes
Absolute relative error (AbsRel)· 2024-03-15
0.089
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Depth EstimationonCityscapes
RMSE· 2024-03-15
5.481
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Depth EstimationonCityscapes
RMSE log· 2024-03-15
0.139
SOTA
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662
Depth EstimationonCityscapes
Square relative error (SqRel)· 2024-03-15
1.051
best: 0.792 (Manydepth2)
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662

Medical1 result

Semantic SegmentationonCityscapes val
mIoU· 2024-03-15
76.41
best: 90.3 (EfficientPS (Cityscapes-fine))
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662

Audio1 result

10-shot image generationonCityscapes val
mIoU· 2024-03-15
76.41
best: 90.3 (EfficientPS (Cityscapes-fine))
SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images arXiv:2403.10662