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/VAT

VAT

Reported on 28 benchmarks across 7 tasks · 2 papers · 17 SOTA

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

Computer Vision20 results

Image MatchingonSPair-71k
PCK· 2021-12-22
54.2
best: 85.6 (GeoAware-SC (Supervised, AP-10K P.T.))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Image MatchingonPF-WILLOW
PCK· 2021-12-22
81
best: 84.3 (LDMCorrespondences)
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot Semantic SegmentationonFSS-1000 (5-shot)
Mean IoU· 2021-12-22
90.6
best: 91.7 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot Semantic SegmentationonFSS-1000 (1-shot)
Mean IoU· 2021-12-22
90
best: 90.8 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot Semantic SegmentationonPASCAL-5i (1-Shot)
Mean IoU· 2021-12-22
67.5
best: 83.2 (SegGPT (ViT))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Semantic correspondenceonSPair-71k
PCK· 2021-12-22
54.2
best: 85.6 (GeoAware-SC (Supervised, AP-10K P.T.))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Semantic correspondenceonPF-WILLOW
PCK· 2021-12-22
81
best: 84.3 (LDMCorrespondences)
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Image Classificationoncifar10, 250 Labels
Percentage correct· 2017-04-13
63.97
best: 93.73 (ReMixMatch)
SOTA
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Image ClassificationonCIFAR-10, 250 Labels
Percentage error· 2017-04-13
36.03
best: 3.47 (SemiOccam)
SOTA
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Semi-Supervised Image Classificationoncifar10, 250 Labels
Percentage correct· 2017-04-13
63.97
best: 93.73 (ReMixMatch)
SOTA
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Semi-Supervised Image ClassificationonCIFAR-10, 250 Labels
Percentage error· 2017-04-13
36.03
best: 3.47 (SemiOccam)
SOTA
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Image MatchingonPF-PASCAL
PCK· 2021-12-22
92.3
best: 95.8 (DINOv2)
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot Semantic SegmentationonPASCAL-5i (5-Shot)
Mean IoU· 2021-12-22
71.6
best: 89.8 (SegGPT (ViT))
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Semantic correspondenceonPF-PASCAL
PCK· 2021-12-22
92.3
best: 95.8 (DINOv2)
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Image ClassificationonCIFAR-10, 4000 Labels
Percentage error· 2017-04-13
11.36
best: 3.96 (SimMatch)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Image ClassificationonSVHN, 1000 labels
Accuracy· 2017-04-13
94.58
best: 97.58 (EnAET)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Image ClassificationonSVHN, 250 Labels
Accuracy· 2017-04-13
91.59
best: 98.04 (ShrinkMatch)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Semi-Supervised Image ClassificationonCIFAR-10, 4000 Labels
Percentage error· 2017-04-13
11.36
best: 3.96 (SimMatch)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Semi-Supervised Image ClassificationonSVHN, 1000 labels
Accuracy· 2017-04-13
94.58
best: 97.58 (EnAET)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976
Semi-Supervised Image ClassificationonSVHN, 250 Labels
Accuracy· 2017-04-13
91.59
best: 98.04 (ShrinkMatch)
Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning arXiv:1704.03976

Methodology8 results

Few-Shot LearningonFSS-1000 (5-shot)
Mean IoU· 2021-12-22
90.6
best: 91.7 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot LearningonFSS-1000 (1-shot)
Mean IoU· 2021-12-22
90
best: 90.8 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot LearningonPASCAL-5i (1-Shot)
Mean IoU· 2021-12-22
67.5
best: 83.2 (SegGPT (ViT))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Meta-LearningonFSS-1000 (5-shot)
Mean IoU· 2021-12-22
90.6
best: 91.7 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Meta-LearningonFSS-1000 (1-shot)
Mean IoU· 2021-12-22
90
best: 90.8 (DACM (ResNet-101))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Meta-LearningonPASCAL-5i (1-Shot)
Mean IoU· 2021-12-22
67.5
best: 83.2 (SegGPT (ViT))
SOTA
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Few-Shot LearningonPASCAL-5i (5-Shot)
Mean IoU· 2021-12-22
71.6
best: 89.8 (SegGPT (ViT))
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685
Meta-LearningonPASCAL-5i (5-Shot)
Mean IoU· 2021-12-22
71.6
best: 89.8 (SegGPT (ViT))
Cost Aggregation Is All You Need for Few-Shot Segmentation arXiv:2112.11685