A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Filters
Class-Balanced Pixel-Level Self-Labeling for Domain Adaptive Semantic Segmentation
[article]
2022
arXiv
pre-print
Preserved Fulltext
The proposed method, namely Class-balanced Pixel-level Self-Labeling (CPSL), improves the segmentation performance on target domain over state-of-the-arts by a large margin, especially on long-tailed categories ...
Domain adaptive semantic segmentation aims to learn a model with the supervision of source domain data, and produce satisfactory dense predictions on unlabeled target domain. ...
[55] proposed a class-balanced self-training method for domain adaption of semantic segmentation. To reduce the noise in pseudo labels, Zou et al. ...
arXiv:2203.09744v1
fatcat:uwpbcc7atfbsllnl4pxk6kla4m
Learning from Scale-Invariant Examples for Domain Adaptation in Semantic Segmentation
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Self-supervised learning approaches for unsupervised domain adaptation (UDA) of semantic segmentation models suffer from challenges of predicting and selecting reasonable good quality pseudo labels. ...
In this paper, we propose a novel approach of exploiting scale-invariance property of the semantic segmentation model for self-supervised domain adaptation. ...
[34] proposed a class-balanced-self-training (CBST) for domain adaptation by generating class-balanced pseudo-labels from images which were assigned labels with most confidence by last state of model ...
arXiv:2007.14449v1
fatcat:xdai7ljwi5dh3jr62m25vlxv54
Source Domain Subset Sampling for Semi-Supervised Domain Adaptation in Semantic Segmentation
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
We propose domain adaptation by sampling and exploiting only a meaningful subset from source data for training. ...
Our key assumption is that the entire source domain data may contain samples that are unhelpful for the adaptation. ...
Our motivation can be applied in domain adaptation of various tasks other than semantic segmentation. However, pixel-level sampling is dependent on semantic segmentation. ...
arXiv:2205.00312v2
fatcat:aaclo4aktzg5nb2tm22pjnecjy
Multi-level Domain Adaptation for Lane Detection
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
To address the issue, we propose the Multi-level Domain Adaptation (MLDA) framework, a new perspective to handle cross-domain lane detection at three complementary semantic levels of pixel, instance and ...
Specifically, at pixel level, we propose to apply cross-class confidence constraints in self-training to tackle the imbalanced confidence distribution of lane and background. ...
In pixel-level adaptation, we propose a cross-domain class balance scheme based on confidence constrained self-training. ...
arXiv:2206.10692v1
fatcat:xze7vjgreffktblya3tbmhr3py
Domain Adaptive Semantic Segmentation without Source Data
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
In positive learning, we select the class-balanced pseudo-labeled pixels with intra-class threshold, while in negative learning, for each pixel, we investigate which category the pixel does not belong ...
Domain adaptive semantic segmentation is recognized as a promising technique to alleviate the domain shift between the labeled source domain and the unlabeled target domain in many real-world applications ...
However, in semantic segmentation, (1) generating a pixel-level image is impractical, (2) and instance-level pseudo-label methods may fail in the dense task, whose objective is a pixel rather than an image ...
arXiv:2110.06484v1
fatcat:dd3kxoqrvnaefahpajeol6pjei
Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training
[article]
2018
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Other Versions
On top of self-training, we also propose a novel class-balanced self-training framework to avoid the gradual dominance of large classes on pseudo-label generation, and introduce spatial priors to refine ...
Unsupervised domain adaptation (UDA) seeks to overcome such problem without target domain labels. ...
As a result, domain adaptation for semantic segmentation recently emerged as a hot topic. ...
arXiv:1810.07911v2
fatcat:hspcyyc3sbcfnpdfth2mirfdxe
Distribution regularized self-supervised learning for domain adaptation of semantic segmentation
2022
Image and Vision Computing
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Note that this fulltext copy is not of the "primary" version of this work. The version it corresponds to is:
2022 pre-print arXiv:2206.09683v1 fatcat:d4ivo5nczjgw3c5uwo6wonzksi
This paper proposes a novel pixel-level distribution regularization scheme (DRSL) for self-supervised domain adaptation of semantic segmentation. ...
Due to the domain shift, this decision boundary is unaligned in the target domain, resulting in noisy pseudo labels adversely affecting self-supervised domain adaptation. ...
For self-supervised domain adaptation, selection of pixels as pseudo-labels is an important step as the adaptation process depends on the quality of pseudo-labels. ...
doi:10.1016/j.imavis.2022.104504
fatcat:dtn6qgbr4jayfecrvm6mzko3sa
ESL: Entropy-guided Self-supervised Learning for Domain Adaptation in Semantic Segmentation
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In particular, self-supervised learning (SSL) has recently become an effective strategy for UDA in semantic segmentation. ...
In addition, producing the extensive pixel-level annotations that the task requires comes at a great cost. ...
To this end, most frameworks incorporate semantic segmentation modules to obtain class-label predictions for all input scene pixels. ...
arXiv:2006.08658v1
fatcat:yakgyozlhneixagkpfs372rovu
MLAN: Multi-Level Adversarial Network for Domain Adaptive Semantic Segmentation
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Recent progresses in domain adaptive semantic segmentation demonstrate the effectiveness of adversarial learning (AL) in unsupervised domain adaptation. ...
In addition, we design a multi-level consistency map that can guide domain adaptation in both input space (i.e., image-to-image translation) and output space (i.e., self-training) effectively. ...
Acknowledgment This research was conducted at Singtel Cognitive and Artificial Intelligence Lab for Enterprises (SCALE@NTU), which is a collaboration between Singapore Telecommunications Limited (Singtel ...
arXiv:2103.12991v2
fatcat:ordpzyl5gnepfhyifjgbrhfk3q
Unsupervised Domain Adaptation for Semantic Segmentation via Class-Balanced Self-training
[chapter]
2018
Lecture Notes in Computer Science
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
Unsupervised domain adaptation (UDA) seeks to overcome such problem without target domain labels. ...
Recent deep networks achieved state of the art performance on a variety of semantic segmentation tasks. ...
-To solve the class imbalance problem of pseudo-labels in ST, we propose a novel class-balanced self-training (CBST) adaptation for semantic segmentation. ...
doi:10.1007/978-3-030-01219-9_18
fatcat:nvhstjstovg3zpgjogn5gqlu74
Unsupervised Domain Adaptation by Optical Flow Augmentation in Semantic Segmentation
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Solving this can totally eliminate the need for labeling real-life datasets completely. Class balanced self-training is one of the existing techniques that attempt to reduce the domain gap. ...
Hence, by augmenting images with dense optical flow map, domain adaptation in semantic segmentation can be improved. ...
Introduction Semantic segmentation involves being able to label every pixel in an image by assigning each pixel to a specific class. The Problem can be poised as a supervised learning one. ...
arXiv:1911.09652v1
fatcat:qq5shbwxpvd2neetnvx45x2qmi
SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
and class-balanced pixel embedding space across domains. ...
Domain adaptive semantic segmentation attempts to make satisfactory dense predictions on an unlabeled target domain by utilizing the model trained on a labeled source domain. ...
Self-training domain adaptation revisit Here, we give an overview over a self-training method [31] for evaluating different semantic-guided pixel contrasts. ...
arXiv:2204.08808v1
fatcat:tfqcpyedsnbnfejfpd2ws4muhe
MLSL: Multi-Level Self-Supervised Learning for Domain Adaptation with Spatially Independent and Semantically Consistent Labeling
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In this paper, we propose a multi-level self-supervised learning model for domain adaptation of semantic segmentation. ...
Image level pseudo weak-labels, PWL, are computed to guide domain adaptation by capturing global context similarity in source and domain at latent space level. ...
A Multi-level self learning strategy for UDA of semantic segmentation by generating pseudo-labels at finegrain pixel-level and image level, helping identify domain invariant features at both latent and ...
arXiv:1909.13776v1
fatcat:bbfmefaekra23eqge6elppmluy
Combining Scale-Invariance and Uncertainty for Self-Supervised Domain Adaptation of Foggy Scenes Segmentation
[article]
2022
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
Other Versions
This paper presents FogAdapt, a novel approach for domain adaptation of semantic segmentation for dense foggy scenes. ...
We propose a self-entropy and multi-scale information augmented self-supervised domain adaptation method (FogAdapt) to minimize the domain shift in foggy scenes segmentation. ...
A self-supervised domain adaptation strategy for foggy scenes segmentation with pixel-level pseudo-labels to adapt the output space. 2. ...
arXiv:2201.02588v2
fatcat:4ggc4aavgjdv3agtifsxns3p5m
Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach
[article]
2019
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We report state-of-the-art results for the adaptation from both GTAV and SYNTHIA to Cityscapes, two popular settings in unsupervised domain adaptation for semantic segmentation. ...
We propose a new approach, called self-motivated pyramid curriculum domain adaptation (PyCDA), to facilitate the adaptation of semantic segmentation neural networks from synthetic source domains to real ...
Conclusion We propose a novel method called self-motivated pyramid curriculum domain adaptation (PyCDA) for pixel-level semantic segmentation. ...
arXiv:1908.09547v1
fatcat:4k54bqcezjfjpbwz3dt4cber6e
« Previous
Showing results 1 — 15 out of 4,540 results