Cluster, Split, Fuse, and Update: Meta-Learning for Open Compound Domain Adaptive Semantic Segmentation

Open compound domain adaptation (OCDA) is a domain adaptation setting, where target domain is modeled as a compound of multiple unknown homogeneous domains, which brings the advantage of improved generalization to unseen domains. In this work, we propose a principled meta-learning based approach to OCDA for semantic segmentation, MOCDA, by modeling the unlabeled target domain continuously. Our approach consists of four key steps. First, we cluster target domain into multiple sub-target domains by image styles, extracted in an unsupervised manner. Then, different sub-target domains are split into independent branches, for which batch normalization parameters are learnt to treat them independently. A meta-learner is thereafter deployed to learn to fuse sub-target domain-specific predictions, conditioned upon the style code. Meanwhile, we learn to online update the model by model-agnostic meta-learning (MAML) algorithm, thus to further improve generalization. We validate the benefits of our approach by extensive experiments on synthetic-to-real knowledge transfer benchmark datasets, where we achieve the state-of-the-art performance in both compound and open domains.

[1]  Zhou Yu,et al.  Domain Adaptive Dialog Generation via Meta Learning , 2019, ACL.

[2]  Luc Van Gool,et al.  Learning Semantic Segmentation From Synthetic Data: A Geometrically Guided Input-Output Adaptation Approach , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Sethuraman Panchanathan,et al.  Deep Hashing Network for Unsupervised Domain Adaptation , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Sergey Levine,et al.  Meta-Learning with Implicit Gradients , 2019, NeurIPS.

[5]  George Papandreou,et al.  Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation , 2018, ECCV.

[6]  In-So Kweon,et al.  Discover, Hallucinate, and Adapt: Open Compound Domain Adaptation for Semantic Segmentation , 2021, NeurIPS.

[7]  Victor S. Lempitsky,et al.  Unsupervised Domain Adaptation by Backpropagation , 2014, ICML.

[8]  Luc Van Gool,et al.  DHP: Differentiable Meta Pruning via HyperNetworks , 2020, ECCV.

[9]  George Papandreou,et al.  Rethinking Atrous Convolution for Semantic Image Segmentation , 2017, ArXiv.

[10]  Jan Kautz,et al.  Multimodal Unsupervised Image-to-Image Translation , 2018, ECCV.

[11]  Alexei A. Efros,et al.  Unbiased look at dataset bias , 2011, CVPR 2011.

[12]  Marc Pollefeys,et al.  Slanted Stixels: Representing San Francisco's Steepest Streets , 2017, BMVC.

[13]  Andreas Geiger,et al.  Augmented Reality Meets Computer Vision: Efficient Data Generation for Urban Driving Scenes , 2017, International Journal of Computer Vision.

[14]  Taesung Park,et al.  CyCADA: Cycle-Consistent Adversarial Domain Adaptation , 2017, ICML.

[15]  Wen Li,et al.  Domain Generalization and Adaptation Using Low Rank Exemplar SVMs , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Patrick Pérez,et al.  ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Bohyung Han,et al.  Domain-Specific Batch Normalization for Unsupervised Domain Adaptation , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[18]  Philip David,et al.  Domain Adaptation for Semantic Segmentation of Urban Scenes , 2017 .

[19]  Daniel C. Castro,et al.  Domain Generalization via Model-Agnostic Learning of Semantic Features , 2019, NeurIPS.

[20]  Michael I. Jordan,et al.  Learning Transferable Features with Deep Adaptation Networks , 2015, ICML.

[21]  Sepp Hochreiter,et al.  Learning to Learn Using Gradient Descent , 2001, ICANN.

[22]  Yinghuan Shi,et al.  Generalizable Semantic Segmentation via Model-agnostic Learning and Target-specific Normalization , 2020, ArXiv.

[23]  Iasonas Kokkinos,et al.  DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[24]  Koby Crammer,et al.  A theory of learning from different domains , 2010, Machine Learning.

[25]  Fengmao Lv,et al.  Constructing Self-Motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[26]  Timothy Hospedales,et al.  Online Meta-Learning for Multi-Source and Semi-Supervised Domain Adaptation , 2020, ECCV.

[27]  Natalia Gimelshein,et al.  PyTorch: An Imperative Style, High-Performance Deep Learning Library , 2019, NeurIPS.

[28]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[29]  Vladlen Koltun,et al.  Playing for Data: Ground Truth from Computer Games , 2016, ECCV.

[30]  D. Tao,et al.  Deep Domain Generalization via Conditional Invariant Adversarial Networks , 2018, ECCV.

[31]  Paul Newman,et al.  1 year, 1000 km: The Oxford RobotCar dataset , 2017, Int. J. Robotics Res..

[32]  Silvio Savarese,et al.  Generalizing to Unseen Domains via Adversarial Data Augmentation , 2018, NeurIPS.

[33]  Philipp Krahenbuhl,et al.  Domain Adaptation Through Task Distillation , 2020, ECCV.

[34]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[35]  Yongxin Yang,et al.  Deeper, Broader and Artier Domain Generalization , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[36]  Trevor Darrell,et al.  Adversarial Discriminative Domain Adaptation , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[37]  Liang Lin,et al.  Blending-Target Domain Adaptation by Adversarial Meta-Adaptation Networks , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[38]  Trevor Darrell,et al.  Fully Convolutional Networks for Semantic Segmentation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[39]  Vladimir Pavlovic,et al.  Unsupervised Multi-Target Domain Adaptation: An Information Theoretic Approach , 2018, IEEE Transactions on Image Processing.

[40]  Alex ChiChung Kot,et al.  Domain Generalization with Adversarial Feature Learning , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[41]  S. P. Lloyd,et al.  Least squares quantization in PCM , 1982, IEEE Trans. Inf. Theory.

[42]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[43]  Yang Zou,et al.  Domain Adaptation for Semantic Segmentation via Class-Balanced Self-Training , 2018, ArXiv.

[44]  Ivor W. Tsang,et al.  Domain Adaptation via Transfer Component Analysis , 2009, IEEE Transactions on Neural Networks.

[45]  Stella X. Yu,et al.  Open Compound Domain Adaptation , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[46]  Dacheng Tao,et al.  Category Anchor-Guided Unsupervised Domain Adaptation for Semantic Segmentation , 2019, NeurIPS.

[47]  Sebastian Ramos,et al.  The Cityscapes Dataset for Semantic Urban Scene Understanding , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[48]  Oliver Zendel,et al.  WildDash - Creating Hazard-Aware Benchmarks , 2018, ECCV.

[49]  Trevor Darrell,et al.  BDD100K: A Diverse Driving Video Database with Scalable Annotation Tooling , 2018, ArXiv.

[50]  Luc Van Gool,et al.  ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[51]  Luc Van Gool,et al.  Exploring Cross-Image Pixel Contrast for Semantic Segmentation , 2021, ArXiv.

[52]  L. Gool,et al.  Learning Discriminative Model Prediction for Tracking , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[53]  Bernhard Schölkopf,et al.  Correcting Sample Selection Bias by Unlabeled Data , 2006, NIPS.

[54]  Luc Van Gool,et al.  DLOW: Domain Flow for Adaptation and Generalization , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[55]  Xi Peng,et al.  Learning to Learn Single Domain Generalization , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[56]  Ming-Hsuan Yang,et al.  Learning to Adapt Structured Output Space for Semantic Segmentation , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[57]  Marcin Andrychowicz,et al.  Learning to learn by gradient descent by gradient descent , 2016, NIPS.

[58]  Kurt Keutzer,et al.  Multi-source Domain Adaptation for Semantic Segmentation , 2019, NeurIPS.

[59]  Luc Van Gool,et al.  Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation , 2020, ECCV.

[60]  Xiaoou Tang,et al.  Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net , 2018, ECCV.

[61]  Swami Sankaranarayanan,et al.  Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[62]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[63]  Michael I. Jordan,et al.  Deep Transfer Learning with Joint Adaptation Networks , 2016, ICML.

[64]  Bo Wang,et al.  Moment Matching for Multi-Source Domain Adaptation , 2018, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[65]  Krzysztof Czarnecki,et al.  Canadian Adverse Driving Conditions dataset , 2020, Int. J. Robotics Res..

[66]  Swami Sankaranarayanan,et al.  MetaReg: Towards Domain Generalization using Meta-Regularization , 2018, NeurIPS.

[67]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[68]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[69]  Tieniu Tan,et al.  Meta-SR: A Magnification-Arbitrary Network for Super-Resolution , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[70]  Swami Sankaranarayanan,et al.  Unsupervised Domain Adaptation for Semantic Segmentation with GANs , 2017, ArXiv.

[71]  Yuan Shi,et al.  Geodesic flow kernel for unsupervised domain adaptation , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[72]  Trevor Darrell,et al.  Adapting Visual Category Models to New Domains , 2010, ECCV.

[73]  Tatsuya Harada,et al.  Maximum Classifier Discrepancy for Unsupervised Domain Adaptation , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[74]  Trevor Darrell,et al.  FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation , 2016, ArXiv.

[75]  Yongxin Yang,et al.  Learning to Generalize: Meta-Learning for Domain Generalization , 2017, AAAI.