Multi-Domain Conditional Image Translation: Translating Driving Datasets from Clear-Weather to Adverse Conditions

Vinod, Vishal; Prabhakar, K. Ram; Babu, R. Venkatesh; Chakraborty, Anirban

doi:10.1109/iccvw54120.2021.00182

Cited by 5 publications

(2 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To efficiently translate images among more than two domains, multi-domain image translation methods [21], [22], [23], e.g., StarGAN [7], build unified models to learn oneto-one mappings between the shared latent space and input domains. To sample multimodal translation results between domains, MUNIT [24] and DRIT [25] decompose the latent space into a shared content space and an unshared style space.…”

Section: Unit Of Multiple Domainsmentioning

confidence: 99%

Domain-Scalable Unpaired Image Translation Via Latent Space Anchoring

Huang

Wei

et al. 2023

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Unpaired image-to-image translation (UNIT) aims to map images between two visual domains without paired training data. However, given a UNIT model trained on certain domains, it is difficult for current methods to incorporate new domains because they often need to train the full model on both existing and new domains. To address this problem, we propose a new domain-scalable UNIT method, termed as latent space anchoring, which can be efficiently extended to new visual domains and does not need to fine-tune encoders and decoders of existing domains. Our method anchors images of different domains to the same latent space of frozen GANs by learning lightweight encoder and regressor models to reconstruct single-domain images. In the inference phase, the learned encoders and decoders of different domains can be arbitrarily combined to translate images between any two domains without fine-tuning. Experiments on various datasets show that the proposed method achieves superior performance on both standard and domain-scalable UNIT tasks in comparison with the state-of-the-art methods.

show abstract

Section: Unit Of Multiple Domainsmentioning

confidence: 99%

Domain-Scalable Unpaired Image Translation Via Latent Space Anchoring

Huang

Wei

et al. 2023

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…CBST [21] and CRST [27] are two representative self-training strategies for domain adaptive semantic segmentation. Recent methods CuDA-Net [57], FIFO [58] and CMDIT [59] focus on bridging the domain gap between the clear images and foggy images to improve the performance of foggy scene segmentation. FogAdapt+ [60] combines the scale-invariance and uncertainty to minimize the domain shift in foggy scenes segmentation.…”

Section: A Experimental Settings 1) Datasetsmentioning

confidence: 99%

Unsupervised Foggy Scene Understanding via Self Spatial-Temporal Label Diffusion

Liao

Chen

Xiao

et al. 2022

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Understanding foggy image sequence in driving scene is critical for autonomous driving, but it remains a challenging task due to the difficulty in collecting and annotating real-world images of adverse weather. Recently, self-training strategy has been considered as a powerful solution for unsupervised domain adaptation, which iteratively adapts the model from the source domain to the target domain by generating target pseudo labels and re-training the model. However, the selection of confident pseudo labels inevitably suffers from the conflict between sparsity and accuracy, both of which will lead to suboptimal models. To tackle this problem, we exploit the characteristics of the foggy image sequence of driving scenes to densify the confident pseudo labels. Specifically, based on the two discoveries of local spatial similarity and adjacent temporal correspondence of the sequential image data, we propose a novel Target-Domain driven pseudo label Diffusion (TDo-Dif) scheme. It employs superpixels and optical flows to identify the spatial similarity and temporal correspondence, respectively, and then diffuses the confident but sparse pseudo labels within a superpixel or a temporal corresponding pair linked by the flow. Moreover, to ensure the feature similarity of the diffused pixels, we introduce local spatial similarity loss and temporal contrastive loss in the model re-training stage. Experimental results show that our TDo-Dif scheme helps the adaptive model achieve 51.92% and 53.84% mean intersection-over-union (mIoU) on two publicly available natural foggy datasets (Foggy Zurich and Foggy Driving), which exceeds the state-of-the-art unsupervised domain adaptive semantic segmentation methods. The proposed method can also be applied to non-sequential images in the target domain by considering only spatial similarity. Models and data can be found at https://github.com/velor2012/TDo-Dif.

show abstract

SST-GAN: Single Sample-based Realistic Traffic Image Generation for Parallel Vision

Wang

Tian

et al. 2022

2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC)

View full text Add to dashboard Cite

Multi-Domain Conditional Image Translation: Translating Driving Datasets from Clear-Weather to Adverse Conditions

Cited by 5 publications

References 57 publications

Domain-Scalable Unpaired Image Translation Via Latent Space Anchoring

Domain-Scalable Unpaired Image Translation Via Latent Space Anchoring

Unsupervised Foggy Scene Understanding via Self Spatial-Temporal Label Diffusion

SST-GAN: Single Sample-based Realistic Traffic Image Generation for Parallel Vision

Contact Info

Product

Resources

About