Visualizing Adapted Knowledge in Domain Transfer

Hou, Yunzhong; Zheng, Liang

doi:10.1109/cvpr46437.2021.01361

Cited by 52 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One line is based on sample generation strategy. For example, 3C-GAN [21] and SDDA [19] generate labeled samples with target domain style for training; VDM-DA [30] generates source domain style features then aligns the generated features with target features; SFIT [11] utilizes the batch-norm layers of the source model to generate images with source domain style and aligns the output predictions. Another line uses pseudo label strategy.…”

Section: Source-free Domain Adaptationmentioning

confidence: 99%

Source-Free Object Detection by Learning to Overlook Domain Style

Zhu

et al. 2022

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

Source-free object detection (SFOD) needs to adapt a detector pre-trained on a labeled source domain to a target domain, with only unlabeled training data from the target domain. Existing SFOD methods typically adopt the pseudo labeling paradigm with model adaption alternating between predicting pseudo labels and fine-tuning the model. This approach suffers from both unsatisfactory accuracy of pseudo labels due to the presence of domain shift and limited use of target domain training data. In this work, we present a novel Learning to Overlook Domain Style (LODS) method with such limitations solved in a principled manner. Our idea is to reduce the domain shift effect by enforcing the model to overlook the target domain style, such that model adaptation is simplified and becomes easier to carry on. To that end, we enhance the style of each target domain image and leverage the style degree difference between the original image and the enhanced image as a self-supervised signal for model adaptation. By treating the enhanced image as an auxiliary view, we exploit a studentteacher architecture for learning to overlook the style degree difference against the original image, also characterized with a novel style enhancement algorithm and graph alignment constraint. Extensive experiments demonstrate that our LODS yields new state-of-the-art performance on four benchmarks.

show abstract

Section: Source-free Domain Adaptationmentioning

confidence: 99%

Source-Free Object Detection by Learning to Overlook Domain Style

Zhu

et al. 2022

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

show abstract

“…CPGA [42] generates source avatar prototypes via contrastive learning and achieves adaptation with target pseudo labels. SFIT [16] designs a two-branch framework to achieve image translation and fine-tunes the target model with the generated images. However, generationbased methods always require additional models and are difficult to train.…”

Section: Model Adaptationmentioning

confidence: 99%

AdaptGuard: Defending Against Universal Attacks for Model Adaptation

Sheng¹,

Ji²,

He³

et al. 2023

Preprint

View full text Add to dashboard Cite

Model adaptation aims at solving the domain transfer problem under the constraint of only accessing the pretrained source models. With the increasing considerations of data privacy and transmission efficiency, this paradigm has been gaining recent popularity. This paper studies the vulnerability to universal attacks transferred from the source domain during model adaptation algorithms due to the existence of the malicious providers. We explore both universal adversarial perturbations and backdoor attacks as loopholes on the source side and discover that they still survive in the target models after adaptation. To address this issue, we propose a model preprocessing framework, named AdaptGuard, to improve the security of model adaptation algorithms. AdaptGuard avoids direct use of the risky source parameters through knowledge distillation and utilizes the pseudo adversarial samples under adjusted radius to enhance the robustness. AdaptGuard is a plug-and-play module that requires neither robust pretrained models nor any changes for the following model adaptation algorithms. Extensive results on three commonly used datasets and two popular adaptation methods validate that AdaptGuard can effectively defend against universal attacks and maintain clean accuracy in the target domain simultaneously. We hope this research will shed light on the safety and robustness of transfer learning.

show abstract

“…In image style transfer, the use of unsupervised unpaired images has become the most widely used method for style transfer [12] . This type of algorithm only needs to collect similar images from the same domain, without collecting the original and target domains of each image, greatly saving the workload of collecting images.…”

Section: Algorithmmentioning

confidence: 99%

DualGGAN: A New Facial Style Transfer Network

Zhao,

Peng,

Zhang

2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

In response to the background penetration problem of unsupervised style transfer algorithms in most cases, a Transformer style transfer network DualGGAN based on dual generators and fusion of relative position encoding is proposed. The network is trained using the least squares generative adversarial network, and the neural network is used as the image feature extractor to generate feature maps to obtain facial image features with attention weights from the feature maps, utilizing relative position encoding and mask loss to jointly constrain feature region style transfer. The experimental results show that the DualGGAN network effectively reduces artifact generation when implementing facial style transfer, maintains good background consistency, and has good generalization ability. Experiments have shown that the FID and KID indicators on the cat2dog and older2adult datasets are significantly improved compared to other algorithms.

show abstract

Visualizing Adapted Knowledge in Domain Transfer

Cited by 52 publications

References 19 publications

Source-Free Object Detection by Learning to Overlook Domain Style

Source-Free Object Detection by Learning to Overlook Domain Style

AdaptGuard: Defending Against Universal Attacks for Model Adaptation

DualGGAN: A New Facial Style Transfer Network

Contact Info

Product

Resources

About