2022-10-21 更新
Intra-Source Style Augmentation for Improved Domain Generalization
Authors:Yumeng Li, Dan Zhang, Margret Keuper, Anna Khoreva
The generalization with respect to domain shifts, as they frequently appear in applications such as autonomous driving, is one of the remaining big challenges for deep learning models. Therefore, we propose an intra-source style augmentation (ISSA) method to improve domain generalization in semantic segmentation. Our method is based on a novel masked noise encoder for StyleGAN2 inversion. The model learns to faithfully reconstruct the image preserving its semantic layout through noise prediction. Random masking of the estimated noise enables the style mixing capability of our model, i.e. it allows to alter the global appearance without affecting the semantic layout of an image. Using the proposed masked noise encoder to randomize style and content combinations in the training set, ISSA effectively increases the diversity of training data and reduces spurious correlation. As a result, we achieve up to $12.4\%$ mIoU improvements on driving-scene semantic segmentation under different types of data shifts, i.e., changing geographic locations, adverse weather conditions, and day to night. ISSA is model-agnostic and straightforwardly applicable with CNNs and Transformers. It is also complementary to other domain generalization techniques, e.g., it improves the recent state-of-the-art solution RobustNet by $3\%$ mIoU in Cityscapes to Dark Z\”urich.
PDF Accepted at WACV 2023
点此查看论文截图
How to Boost Face Recognition with StyleGAN?
Authors:Artem Sevastopolsky, Yury Malkov, Nikita Durasov, Luisa Verdoliva, Matthias Nießner
State-of-the-art face recognition systems require huge amounts of labeled training data. Given the priority of privacy in face recognition applications, the data is limited to celebrity web crawls, which have issues such as skewed distributions of ethnicities and limited numbers of identities. On the other hand, the self-supervised revolution in the industry motivates research on adaptation of the related techniques to facial recognition. One of the most popular practical tricks is to augment the dataset by the samples drawn from the high-resolution high-fidelity models (e.g. StyleGAN-like), while preserving the identity. We show that a simple approach based on fine-tuning an encoder for StyleGAN allows to improve upon the state-of-the-art facial recognition and performs better compared to training on synthetic face identities. We also collect large-scale unlabeled datasets with controllable ethnic constitution — AfricanFaceSet-5M (5 million images of different people) and AsianFaceSet-3M (3 million images of different people) and we show that pretraining on each of them improves recognition of the respective ethnicities (as well as also others), while combining all unlabeled datasets results in the biggest performance increase. Our self-supervised strategy is the most useful with limited amounts of labeled training data, which can be beneficial for more tailored face recognition tasks and when facing privacy concerns. Evaluation is provided based on a standard RFW dataset and a new large-scale RB-WebFace benchmark.
PDF
点此查看论文截图
Infrared and visible image fusion via dual-domain adversarial learning
Authors:Xiaowen Liu, Renhua Wang, Hongtao Huo, Jing Li, Xin Yang
The GAN-based infrared and visible image fusion methods have gained ever-increasing attention due to its effectiveness and superiority. However, the existing methods adopt the global pixel distribution of source images as the basis for discrimination, which fails to focus on the key modality information. Moreover, the dual-discriminator based methods suffer from the confrontation between the discriminators. To this end, we propose a dual-domain adversarial based infrared and visible image fusion method (D2AFGAN). In this method, two unique discrimination strategies are designed to improve the fusion performance. Specifically, we introduce the spatial attention modules (SAM) into the generator to obtain the spatial attention maps, and then the attention maps are utilized to force the discrimination of infrared images to focus on the target regions. In addition, we extend the discrimination range of visible information to the wavelet subspace, which can force the generator to restore the high-frequency details of visible images. Ablation experiments demonstrate the effectiveness of our method in eliminating the confrontation between discriminators. And the comparison experiments on public datasets demonstrate the effectiveness and superiority of the proposed method.
PDF
点此查看论文截图
PalGAN: Image Colorization with Palette Generative Adversarial Networks
Authors:Yi Wang, Menghan Xia, Lu Qi, Jing Shao, Yu Qiao
Multimodal ambiguity and color bleeding remain challenging in colorization. To tackle these problems, we propose a new GAN-based colorization approach PalGAN, integrated with palette estimation and chromatic attention. To circumvent the multimodality issue, we present a new colorization formulation that estimates a probabilistic palette from the input gray image first, then conducts color assignment conditioned on the palette through a generative model. Further, we handle color bleeding with chromatic attention. It studies color affinities by considering both semantic and intensity correlation. In extensive experiments, PalGAN outperforms state-of-the-arts in quantitative evaluation and visual comparison, delivering notable diverse, contrastive, and edge-preserving appearances. With the palette design, our method enables color transfer between images even with irrelevant contexts.
PDF Accepted at ECCV 2022
点此查看论文截图