DLOW: Domain Flow for Adaptation and Generalization

By Rui Gong, Wen Li, Yuhua Chen, Luc Van Gool. 2018

400 citations recorded by the source corpus.

In this work, we present a domain flow generation(DLOW) model to bridge two different domains by generating a continuous sequence of intermediate domains flowing from one domain to the other. The benefits of our DLOW model are two-fold. First, it is able to transfer source images into different styles in the intermediate domains. The transferred images smoothly bridge the gap between source and target domains, thus easing the domain adaptation task. Second, when multiple target domains are provided for training, our DLOW model is also able to generate new styles of images that are unseen in the training data. We implement our DLOW model based on CycleGAN. A domainness variable is introduced to guide the model to generate the desired intermediate domain images. In the inference phase, a flow of various styles of images can be obtained by varying the domainness variable. We demonstrate the effectiveness of our model for both cross-domain semantic segmentation and the style generalization tasks on benchmark datasets. Our implementation is available at https://github.com/ETHRuiGong/DLOW .

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Prototype-Guided Continual Adaptation for Class-Incremental Unsupervised Domain Adaptation (2022, background): “Step 2 [10,11,12,13,14,15,16,17,18,19] Kettle, Folder, Keyboard, Flipflops, Pencil, Bed, Hammer, ToothBrush, Couch, Bike”
Gradually Vanishing Bridge for Adversarial Domain Adaptation (2020, methodology): “In deep methods, DSN [2], DISE [3] and DLOW [11] formulate the bridge between source and target by reconstructing input images on pixel level.”
What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation (2020, background): “[18] presents the domain flow translation to explore expected intermediate domain.”
Generalizing to Unseen Domains: A Survey on Domain Generalization (2021, methodology): “Some works also study DG for semantic segmentation [79], action”
Learning Semantic Segmentation From Synthetic Data: A Geometrically Guided Input-Output Adaptation Approach (2018, background): “Moreover, inspired by the success of generative adversarial network [44, 17] and image translation techniques [60, 22], a few works have also suggested to transform synthetic images to realistic style, thus reducing domain gap on raw-pixel level [51, 20, 41, 20, 48, 16] and to boost the semantic seg”
Towards Out-Of-Distribution Generalization: A Survey (2021, neutral): “…( Section 3.2 ) [34], [35], [36](Section 3.2) SupervisedLearningModels ( Section 4 ) ( Section 4.1 ) [7], [8], [9], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49] [37], [39], [50], [51], [51], [52], [53], [53], [54], [54], [55], [56], [57], [58], [59], [60] [6],…”
BAPA-Net: Boundary Adaptation and Prototype Alignment for Cross-domain Semantic Segmentation (2021, methodology): “[19] proposes a progressive adaptation method to alleviate the domain shift by controlling the degree of style translation from the source domain to the target domain.”
RobustNet: Improving Domain Generalization in Urban-Scene Segmentation via Instance Selective Whitening (2021, background): “Various approaches such as meta-learning [27, 2, 28, 33], adversarial training [29, 32, 46], autoencoder [15, 29], metric learning [10, 39], data augmentation [64, 16, 67] have been proposed to learn domain-agnostic feature representations.”
Phase Consistent Ecological Domain Adaptation (2020, background): “[16] generates a sequence of intermediate shifted domains from source to target to further improve the transferability by providing multi-style translations.”
FDA: Fourier Domain Adaptation for Semantic Segmentation (2020, methodology): “[17] generates intermediate style images between source and target.”
Gradient Regularized Contrastive Learning for Continual Domain Adaptation (2020, methodology): “The approaches (Mancini et al. 2019; Gong et al. 2019; Wulfmeier, Bewley, and Posner 2018; Hoffman et al. 2014) do not explicitly target the catastrophic forgetting problem, while the approaches (Bobu et al.”
Cross-Domain First Person Audio-Visual Action Recognition through Relative Norm Alignment (2021, background): “Still, another approach is the generative-based one, which operates by performing style-transfer directly on input data [22, 23].”
Confidence Regularized Self-Training (2019, methodology): “Examples of these methods include maximum mean discrepancy (MMD) [33, 62], deep correlation alignment (CORAL) [56], sliced Wasserstein discrepancy [28], adversarial learning at input-level [16, 24, 68], feature level [8, 14, 23, 31, 50, 61, 64], output space level [60], and a variety of follow up wo”
Deep Domain-Adversarial Image Generation for Domain Generalisation (2020, methodology): “A more economical solution is to use unsupervised domain adaptation (UDA) methods (Long et al. 2015; Ganin and Lempitsky 2015; Hoffman et al. 2018; Gong et al. 2019; Chen et al. 2019), which only use unlabelled target data.”
Learning to Adapt to Evolving Domains (2020, background): “Adapting to an intermediate target domain before distant target domain performs better than adapting to them as a whole target [9].”
Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation (2021, background): “As suggested by the theoretical analysis [3], domain alignment methods focus on reducing the distribution mismatch by optimizing some divergence [30, 35] or adopting adversarial training [20, 40] at either the image level [1, 12, 19, 26, 49, 63, 68], the intermediate feature level [6, 27, 48, 58] or”
Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replay (2020, background): “, 2019) attempt to solve a specific scenario in continuous domain adaptation, where no target data is available, but with metadata provided for all domains; (Gong et al., 2019) propose to bridge two domains by generating a continuous flow of intermediate domains between the two original domains; (Ho”
Learning Texture Invariant Representation for Domain Adaptation of Semantic Segmentation (2020, methodology): “Results show model trained on only stylized source dataset outperforms other pixel-level [27, 10] method.”
FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Space (2021, background): “Domain generalization [5, 9, 12, 14, 43, 47, 58, 59] aims to learn a model from multiple source domains such that it can directly generalize to unseen target domains.”