2022-08-04 更新
Paired Cross-Modal Data Augmentation for Fine-Grained Image-to-Text Retrieval
Authors:Hao Wang, Guosheng Lin, Steven C. H. Hoi, Chunyan Miao
This paper investigates an open research problem of generating text-image pairs to improve the training of fine-grained image-to-text cross-modal retrieval task, and proposes a novel framework for paired data augmentation by uncovering the hidden semantic information of StyleGAN2 model. Specifically, we first train a StyleGAN2 model on the given dataset. We then project the real images back to the latent space of StyleGAN2 to obtain the latent codes. To make the generated images manipulatable, we further introduce a latent space alignment module to learn the alignment between StyleGAN2 latent codes and the corresponding textual caption features. When we do online paired data augmentation, we first generate augmented text through random token replacement, then pass the augmented text into the latent space alignment module to output the latent codes, which are finally fed to StyleGAN2 to generate the augmented images. We evaluate the efficacy of our augmented data approach on two public cross-modal retrieval datasets, in which the promising experimental results demonstrate the augmented text-image pair data can be trained together with the original data to boost the image-to-text cross-modal retrieval performance.
PDF Accepted at ACM MM 2022
点此查看论文截图
AvatarGen: a 3D Generative Model for Animatable Human Avatars
Authors:Jianfeng Zhang, Zihang Jiang, Dingdong Yang, Hongyi Xu, Yichun Shi, Guoxian Song, Zhongcong Xu, Xinchao Wang, Jiashi Feng
Unsupervised generation of clothed virtual humans with various appearance and animatable poses is important for creating 3D human avatars and other AR/VR applications. Existing methods are either limited to rigid object modeling, or not generative and thus unable to synthesize high-quality virtual humans and animate them. In this work, we propose AvatarGen, the first method that enables not only non-rigid human generation with diverse appearance but also full control over poses and viewpoints, while only requiring 2D images for training. Specifically, it extends the recent 3D GANs to clothed human generation by utilizing a coarse human body model as a proxy to warp the observation space into a standard avatar under a canonical space. To model non-rigid dynamics, it introduces a deformation network to learn pose-dependent deformations in the canonical space. To improve geometry quality of the generated human avatars, it leverages signed distance field as geometric representation, which allows more direct regularization from the body model on the geometry learning. Benefiting from these designs, our method can generate animatable human avatars with high-quality appearance and geometry modeling, significantly outperforming previous 3D GANs. Furthermore, it is competent for many applications, e.g., single-view reconstruction, reanimation, and text-guided synthesis. Code and pre-trained model will be available.
PDF First two authors contributed equally. Code will be available at https://github.com/jfzhang95/AvatarGen
点此查看论文截图
Backdoor Attack is a Devil in Federated GAN-based Medical Image Synthesis
Authors:Ruinan Jin, Xiaoxiao Li
Deep Learning-based image synthesis techniques have been applied in healthcare research for generating medical images to support open research. Training generative adversarial neural networks (GAN) usually requires large amounts of training data. Federated learning (FL) provides a way of training a central model using distributed data from different medical institutions while keeping raw data locally. However, FL is vulnerable to backdoor attack, an adversarial by poisoning training data, given the central server cannot access the original data directly. Most backdoor attack strategies focus on classification models and centralized domains. In this study, we propose a way of attacking federated GAN (FedGAN) by treating the discriminator with a commonly used data poisoning strategy in backdoor attack classification models. We demonstrate that adding a small trigger with size less than 0.5 percent of the original image size can corrupt the FL-GAN model. Based on the proposed attack, we provide two effective defense strategies: global malicious detection and local training regularization. We show that combining the two defense strategies yields a robust medical image generation.
PDF 13 pages, 4 figures, Accepted by MICCAI 2022 SASHIMI Workshop
点此查看论文截图
Multimodal Controller for Generative Models
Authors:Enmao Diao, Jie Ding, Vahid Tarokh
Class-conditional generative models are crucial tools for data generation from user-specified class labels. Existing approaches for class-conditional generative models require nontrivial modifications of backbone generative architectures to model conditional information fed into the model. This paper introduces a plug-and-play module named `multimodal controller’ to generate multimodal data without introducing additional learning parameters. In the absence of the controllers, our model reduces to non-conditional generative models. We test the efficacy of multimodal controllers on CIFAR10, COIL100, and Omniglot benchmark datasets. We demonstrate that multimodal controlled generative models (including VAE, PixelCNN, Glow, and GAN) can generate class-conditional images of significantly better quality when compared with conditional generative models. Moreover, we show that multimodal controlled models can also create novel modalities of images.
PDF
点此查看论文截图
Generative Adversarial Method Based On Neural Tangent Kernels
Authors:Yu-Rong Zhang, Sheng Yen Chou, Shan-Hung Wu
The recent development of Generative adversarial networks (GANs) has driven many computer vision applications. Despite the great synthesis quality, training GANs often confronts several issues, including non-convergence, mode collapse, and gradient vanishing. There exist several workarounds, for example, regularizing Lipschitz continuity and adopting Wasserstein distance. Although these methods can partially solve the problems, we argue that the problems are result from modeling the discriminator with deep neural networks. In this paper, we base on newly derived deep neural network theories called Neural Tangent Kernel (NTK) and propose a new generative algorithm called generative adversarial NTK (GA-NTK). The GA-NTK models the discriminator as a Gaussian Process (GP). With the help of the NTK theories, the training dynamics of GA-NTK can be described with a closed-form formula. To synthesize data with the closed-form formula, the objectives can be simplified into a single-level adversarial optimization problem. We conduct extensive experiments on real-world datasets, and the results show that GA-NTK can generate images comparable to those by GANs but is much easier to train under various conditions. We also study the current limitations of GA-NTK and propose some workarounds to make GA-NTK more practical.
PDF
点此查看论文截图
Image Augmentation for Satellite Images
Authors:Oluwadara Adedeji, Peter Owoade, Opeyemi Ajayi, Olayiwola Arowolo
This study proposes the use of generative models (GANs) for augmenting the EuroSAT dataset for the Land Use and Land Cover (LULC) Classification task. We used DCGAN and WGAN-GP to generate images for each class in the dataset. We then explored the effect of augmenting the original dataset by about 10% in each case on model performance. The choice of GAN architecture seems to have no apparent effect on the model performance. However, a combination of geometric augmentation and GAN-generated images improved baseline results. Our study shows that GANs augmentation can improve the generalizability of deep classification models on satellite images.
PDF 14 pages, 4 figures, 6 tables. Research project for Introduction to Deep Learning (11785) at Carnegie Mellon University
点此查看论文截图
2022-08-04 更新
Unsupervised Discovery of Semantic Concepts in Satellite Imagery with Style-based Wavelet-driven Generative Models
Authors:Nikos Kostagiolas, Mihalis A. Nicolaou, Yannis Panagakis
In recent years, considerable advancements have been made in the area of Generative Adversarial Networks (GANs), particularly with the advent of style-based architectures that address many key shortcomings - both in terms of modeling capabilities and network interpretability. Despite these improvements, the adoption of such approaches in the domain of satellite imagery is not straightforward. Typical vision datasets used in generative tasks are well-aligned and annotated, and exhibit limited variability. In contrast, satellite imagery exhibits great spatial and spectral variability, wide presence of fine, high-frequency details, while the tedious nature of annotating satellite imagery leads to annotation scarcity - further motivating developments in unsupervised learning. In this light, we present the first pre-trained style- and wavelet-based GAN model that can readily synthesize a wide gamut of realistic satellite images in a variety of settings and conditions - while also preserving high-frequency information. Furthermore, we show that by analyzing the intermediate activations of our network, one can discover a multitude of interpretable semantic directions that facilitate the guided synthesis of satellite images in terms of high-level concepts (e.g., urbanization) without using any form of supervision. Via a set of qualitative and quantitative experiments we demonstrate the efficacy of our framework, in terms of suitability for downstream tasks (e.g., data augmentation), quality of synthetic imagery, as well as generalization capabilities to unseen datasets.
PDF 11 pages, 5 figures, accepted at SETN 2022
点此查看论文截图