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/µ2Net+ (ViT-L/16)

µ2Net+ (ViT-L/16)

Reported on 23 benchmarks across 9 tasks · 1 paper · 15 SOTA

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

Computer Vision19 results

Scene ClassificationonUC Merced Land Use Dataset
Accuracy (%)· 2022-09-15
100
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonDTD
Accuracy· uses extra data· 2022-09-15
82.23
best: 90 (Linear FT(ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image Classificationoncats_vs_dogs
Accuracy· 2022-09-15
99.83
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationoniNaturalist 2018
Top-1 Accuracy· 2022-09-15
80.97
best: 94.6 (OmniVec2)
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonCARS196
Accuracy· 2022-09-15
87.18
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonStanford Online Products
Accuracy· 2022-09-15
89.47
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonSTL-10
Percentage correct· uses extra data· 2022-09-15
99.64
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonEuroSAT
Accuracy (%)· 2022-09-15
99.22
best: 99.41 (DeepEnsembling)
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonImagenette
Accuracy· 2022-09-15
100
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonImageNet-Sketch
Accuracy· 2022-09-15
88.6
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonImageNet-LT
Top-1 Accuracy· 2022-09-15
82.5
best: 82.9 (LIFT (ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Few-Shot Image ClassificationonImageNet-LT
Top-1 Accuracy· 2022-09-15
82.5
best: 82.9 (LIFT (ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonEMNIST-Letters
Accuracy· 2022-09-15
95.03
best: 95.96 (WaveMixLite-112/16)
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonPlaces365
Top 1 Accuracy· uses extra data· 2022-09-15
59.15
best: 65.1 (OmniVec2)
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonOxford-IIIT Pets
Accuracy· 2022-09-15
95.5
best: 97.1 (EffNet-L2 (SAM))
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Image ClassificationonFood-101
Accuracy· 2022-09-15
91.47
best: 98.6 (CAP)
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Fine-Grained Image ClassificationonOxford-IIIT Pets
Accuracy· 2022-09-15
95.5
best: 97.1 (EffNet-L2 (SAM))
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Fine-Grained Image ClassificationonFood-101
Accuracy· 2022-09-15
91.47
best: 98.6 (CAP)
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Domain GeneralizationonImageNet-A
Top-1 accuracy %· uses extra data· 2022-09-15
84.53
best: 94.17 (Model soups (BASIC-L))
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326

Methodology4 results

Generalized Few-Shot ClassificationonImageNet-LT
Top-1 Accuracy· 2022-09-15
82.5
best: 82.9 (LIFT (ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Long-tail LearningonImageNet-LT
Top-1 Accuracy· 2022-09-15
82.5
best: 82.9 (LIFT (ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Generalized Few-Shot LearningonImageNet-LT
Top-1 Accuracy· 2022-09-15
82.5
best: 82.9 (LIFT (ViT-L/14))
SOTA
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326
Domain AdaptationonImageNet-A
Top-1 accuracy %· uses extra data· 2022-09-15
84.53
best: 94.17 (Model soups (BASIC-L))
A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems arXiv:2209.07326