Image Generation on SYNTHIA-to-Cityscapes

Metric: mIoU (13 classes) (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	mIoU (13 classes)▼	Extra Data	Paper	Date↕	Code
1	HRDA + PiPa	74.8	No	PiPa: Pixel- and Patch-wise Self-supervised Lear...	2022-11-14	Code
2	MIC	74	No	MIC: Masked Image Consistency for Context-Enhanc...	2022-12-02	Code
3	HRDA	72.4	No	HRDA: Context-Aware High-Resolution Domain-Adapt...	2022-04-27	Code
4	SePiCo	71.4	No	SePiCo: Semantic-Guided Pixel Contrast for Domai...	2022-04-19	Code
5	CAMix (w DAFormer)	69.2	No	Context-Aware Mixup for Domain Adaptive Semantic...	2021-08-08	Code
6	DAFormer + ProCST	68.2	Yes	ProCST: Boosting Semantic Segmentation Using Pro...	2022-04-25	Code
7	DAFormer	67.4	No	DAFormer: Improving Network Architectures and Tr...	2021-11-29	Code
8	TransDA-B	66.3	No	Smoothing Matters: Momentum Transformer for Doma...	2022-03-15	Code
9	CPSL	65.3	No	Class-Balanced Pixel-Level Self-Labeling for Dom...	2022-03-18	Code
10	ProDA+CRA	63.7	No	Cross-Region Domain Adaptation for Class-level A...	2021-09-14	-
11	ProDA	62	No	Prototypical Pseudo Label Denoising and Target S...	2021-01-26	Code
12	CAMix (w Deeplabv2 ResNet 101)	59.7	No	Context-Aware Mixup for Domain Adaptive Semantic...	2021-08-08	Code
13	IAST(ResNet-101)	57	No	Instance Adaptive Self-Training for Unsupervised...	2020-08-27	Code
14	PyCDA (ResNet-101)	53.3	No	Constructing Self-motivated Pyramid Curriculums ...	2019-08-26	Code
15	FADA (ResNet-101)	52.5	No	Classes Matter: A Fine-grained Adversarial Appro...	2020-07-17	Code
16	Bidirectional Learning (ResNet-101)	51.4	No	Bidirectional Learning for Domain Adaptation of ...	2019-04-24	Code
17	DADA (ResNet-101)	49.8	No	DADA: Depth-aware Domain Adaptation in Semantic ...	2019-04-03	Code
18	LRENT (DeepLabv2)	48.7	No	Confidence Regularized Self-Training	2019-08-26	Code
19	SWD	48.1	No	Sliced Wasserstein Discrepancy for Unsupervised ...	2019-03-10	Code
20	ADVENT	48	No	ADVENT: Adversarial Entropy Minimization for Dom...	2018-11-30	Code
21	Multi-level Adaptation	46.7	No	Learning to Adapt Structured Output Space for Se...	2018-02-28	Code
22	Discriminative Patch (ResNet-101)	46.5	No	Domain Adaptation for Structured Output via Disc...	2019-01-16	Code
23	Single-level Adaptation	45.9	No	Learning to Adapt Structured Output Space for Se...	2018-02-28	Code
24	CAG-UDA	44.5	No	Category Anchor-Guided Unsupervised Domain Adapt...	2019-10-29	Code
25	Domain Invariant Structure Extraction	41.5	No	All about Structure: Adapting Structural Informa...	2019-03-26	Code
26	superpixel + color constancy	29.7	No	A Curriculum Domain Adaptation Approach to the S...	2018-12-24	Code
27	CDA	29	No	Curriculum Domain Adaptation for Semantic Segmen...	2017-07-29	Code
28	FCNs in the wild	20.2	No	FCNs in the Wild: Pixel-level Adversarial and Co...	2016-12-08	Code

#1HRDA + PiPaSOTA
74.8
mIoU (13 classes)· 2022-11-14
PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentation Code
#2MIC
74
mIoU (13 classes)· 2022-12-02
MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation Code
#3HRDASOTA
72.4
mIoU (13 classes)· 2022-04-27
HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation Code
#4SePiCoSOTA
71.4
mIoU (13 classes)· 2022-04-19
SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation Code
#5CAMix (w DAFormer)SOTA
69.2
mIoU (13 classes)· 2021-08-08
Context-Aware Mixup for Domain Adaptive Semantic Segmentation Code
#6DAFormer + ProCST
68.2
mIoU (13 classes)· Extra Data· 2022-04-25
ProCST: Boosting Semantic Segmentation Using Progressive Cyclic Style-Transfer Code
#7DAFormer
67.4
mIoU (13 classes)· 2021-11-29
DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation Code
#8TransDA-B
66.3
mIoU (13 classes)· 2022-03-15
Smoothing Matters: Momentum Transformer for Domain Adaptive Semantic Segmentation Code
#9CPSL
65.3
mIoU (13 classes)· 2022-03-18
Class-Balanced Pixel-Level Self-Labeling for Domain Adaptive Semantic Segmentation Code
#10ProDA+CRA
63.7
mIoU (13 classes)· 2021-09-14
Cross-Region Domain Adaptation for Class-level Alignment
#11ProDASOTA
62
mIoU (13 classes)· 2021-01-26
Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation Code
#12CAMix (w Deeplabv2 ResNet 101)
59.7
mIoU (13 classes)· 2021-08-08
Context-Aware Mixup for Domain Adaptive Semantic Segmentation Code
#13IAST(ResNet-101)SOTA
57
mIoU (13 classes)· 2020-08-27
Instance Adaptive Self-Training for Unsupervised Domain Adaptation Code
#14PyCDA (ResNet-101)SOTA
53.3
mIoU (13 classes)· 2019-08-26
Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach Code
#15FADA (ResNet-101)
52.5
mIoU (13 classes)· 2020-07-17
Classes Matter: A Fine-grained Adversarial Approach to Cross-domain Semantic Segmentation Code
#16Bidirectional Learning (ResNet-101)SOTA
51.4
mIoU (13 classes)· 2019-04-24
Bidirectional Learning for Domain Adaptation of Semantic Segmentation Code
#17DADA (ResNet-101)SOTA
49.8
mIoU (13 classes)· 2019-04-03
DADA: Depth-aware Domain Adaptation in Semantic Segmentation Code
#18LRENT (DeepLabv2)
48.7
mIoU (13 classes)· 2019-08-26
Confidence Regularized Self-Training Code
#19SWDSOTA
48.1
mIoU (13 classes)· 2019-03-10
Sliced Wasserstein Discrepancy for Unsupervised Domain Adaptation Code
#20ADVENTSOTA
48
mIoU (13 classes)· 2018-11-30
ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation Code
#21Multi-level AdaptationSOTA
46.7
mIoU (13 classes)· 2018-02-28
Learning to Adapt Structured Output Space for Semantic Segmentation Code
#22Discriminative Patch (ResNet-101)
46.5
mIoU (13 classes)· 2019-01-16
Domain Adaptation for Structured Output via Discriminative Patch Representations Code
#23Single-level Adaptation
45.9
mIoU (13 classes)· 2018-02-28
Learning to Adapt Structured Output Space for Semantic Segmentation Code
#24CAG-UDA
44.5
mIoU (13 classes)· 2019-10-29
Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentation Code
#25Domain Invariant Structure Extraction
41.5
mIoU (13 classes)· 2019-03-26
All about Structure: Adapting Structural Information across Domains for Boosting Semantic Segmentation Code
#26superpixel + color constancy
29.7
mIoU (13 classes)· 2018-12-24
A Curriculum Domain Adaptation Approach to the Semantic Segmentation of Urban Scenes Code
#27CDASOTA
29
mIoU (13 classes)· 2017-07-29
Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes Code
#28FCNs in the wildSOTA
20.2
mIoU (13 classes)· 2016-12-08
FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation Code