Mutual information maximizing GAN inversion for real face with identity preservation

Recent generative adversarial networks (GANs) have yielded remarkable performance in face image synthesis. GAN inversion embeds an image into the latent space of a pretrained generator, enabling it to be used for real face manipulation. However, current inversion approaches for real faces suffer the dilemma of initialization collapse and identity loss. In this paper, we propose a hierarchical GAN inversion for real faces with identity preservation based on mutual information maximization. We first use a facial domain guaranteed initialization to avoid the initialization collapse. Furthermore, we prove that maximizing the mutual information between inverted faces and their identities is equivalent to minimizing the distance between identity features from inverted and original faces. Optimization for real face inversion with identity preservation is implemented on this mutual information-maximizing constraint. Extensive experimental results show that our approach outperforms state-of-the-art solutions for inverting and editing real faces, particularly in terms of face identity preservation.     

Neural Emotion Detection via Personal Attributes

There has been a recent line of work to automatically detect the emotions of posts in social media. In literature, studies treat posts independently and detect their emotions separately. Different from previous studies, we explore the dependence among relevant posts via authors' backgrounds, since the authors with similar backgrounds, e.g., "gender", "location", tend to express similar emotions. However, personal attributes are not easy to obtain in most social media websites. Accordingly, we propose two approaches to determine personal attributes and capture personal attributes between different posts for emotion detection: the Joint Model with Personal Attention Mechanism (JPA) model is used to detect emotion and personal attributes jointly, and capture the attributes-aware words to connect similar people; the Neural Personal Discrimination (NPD) model is employed to determine the personal attributes from posts and connect the relevant posts with similar attributes for emotion detection. Experimental results show the usefulness of personal attributes in emotion detection, and the effectiveness of the proposed JPA and NPD approaches in capturing personal attributes over the state-of-the-art statistic and neural models.     

基于条件信息卷积生成对抗网络的图像识别

传统的图像识别方法需要大量有标签样本进行训练，且模型训练难以达到稳定。针对这些问题，结合条件生成网络和信息最大化生成网络的结构优势建立了条件信息卷积生成网络（C-Info-DCGAN）。模型增加图像的类别信息和潜在信息作为输入数据，然后利用Q网络去更好地发挥类别信息和潜在信息对训练的引导作用，并且利用深度卷积网络来加强对图像特征的提取能力。实验结果表明，该方法能够加快模型训练收敛速度，并有效提高图像识别的准确率。     

基于改进CGANs的入侵检测方法研究

近年来,机器学习算法在入侵检测系统(IDS)中的应用获得越来越多的关注。然而,传统的机器学习算法更多的依赖于已知样本,因此需要尽可能多的数据样本来对模型进行训练。遗憾地是,随着越来越多未知攻击的出现,且用于训练的攻击样本具有不平衡性,传统的机器学习模型会遇到瓶颈。文章提出一种将改进后的条件生成对抗网络(CGANs)与深度神经网络(DNN)相结合的入侵检测模型(CGANs-DNN),通过解决样本不平衡性问题来提高检测模型对未知攻击类型或只有少数攻击样本类型的检测率。深度神经网络(DNN)具有表征数据潜在特征的能力,而经过改进后的条件CGANs,能够通过学习已知攻击样本潜在数据特征分布,来根据指定类型生成新的攻击样本。此外,与生成对抗网络(GANs)和变分自编码器(VAE)等无监督生成模型相比,CGANsDNN经过改进后加入梯度惩罚项,在训练的稳定性上有了很大地提升。通过NSL-KDD数据集对模型进行评估,与传统算法相比CGANs-DNN不仅在整体准确率、召回率和误报率等方面有更好的性能,而且对未知攻击和只有少数样本的攻击类型具有较高的检测率。     

基于循环一致性对抗网络的室内火焰图像场景迁移

基于深度学习的视频火灾探测模型的训练依赖于大量的正负样本数据,即火灾视频和带有干扰的场景视频。由于很多室内场合禁止点火,导致该场景下的火灾视频样本不足。本文基于生成对抗网络,将其他相似场景下录制的火焰迁移到指定场景,以此增广限制性场合下的火灾视频数据。文中提出将火焰内核预先植入场景使之具备完整的内容信息,再通过添加烟雾和地面反射等风格信息,完成场景与火焰的融合。该方法克服了现有多模态图像转换方法在图像转换过程中因丢失信息而造成的背景失真问题。同时为减少数据采集工作量,采用循环一致性生成对抗网络以解除训练图像必须严格匹配的限制。实验表明,与现有多模态图像转换相比,本文方法可以保证场景中火焰形态的多样性,迁移后的场景具有较高的视觉真实性,所得结果的FID与LPIPS值最小,分别为119.6和0.134 2。     

A framework for improving the efficiency of operational testing through Bayesian adaptive design

When developing a product, it is important to consider product performance from a user perspective. This type of evaluation can be done through operational testing—assessing the ability of representative users to satisfactorily accomplish tasks with the product in operationally representative environments. This process can be expensive and time-consuming, but is critical to understanding whether the product can adequately do the job for which it was designed. We show how an existing design of experiments (DOEs) process for operational testing can be leveraged to construct a Bayesian adaptive design. This design, nested within the larger design created by the DOE process, allows interim analyses to stop testing early for success or futility. Representative simulations are presented to demonstrate how these interim analyses can be used in an operational test setting, and reductions in necessary test events are shown. The application of Bayesian-adaptive design methods will allow future operational testing to be conducted in less time and at less expense, on average, without compromising the ability of the existing process to verify the product meets the user's needs.     

A new VAE-GAN model to synthesize arterial spin labeling images from structural MRI

Arterial spin labeling (ASL) is a relatively new MRI technique that can measure cerebral blood flow, which is of great importance for the diagnosis of dementia diseases. Besides, this valuable imaging modality does not need exogenous tracers and has no radiation, which makes it favorable for elder patients. However, ASL data does lack in many contemporary image-based dementia diseases datasets, which include popular ADNI-1/GO/2/3 datasets. In order to supplement the valuable ASL data, a new Generative adversarial network (GAN)-based model is proposed to synthesize ASL images in this study. This new model is unique, as the popular variational auto-encoder (VAE) has been utilized as the generator of the GAN-based model. Hence, a new VAE-GAN architecture is introduced in this study. In order to demonstrate its superiority, dozens of experiments have been conducted. Experimental results demonstrate that, this new VAE-GAN model is superior to other state-of-the-art ASL image synthesis methods, and the accuracy improvement after incorporating synthesized ASL images from the new model can be as high as 42.41% in dementia diagnosis tasks.     

Peptide-Mimicking Poly(2-oxazoline)s Displaying Potent Antimicrobial Properties

Poly(2-oxazoline)s have excellent biocompatibility and have been used as FDA-approved indirect food additives. The inert property of the hydrophilic poly(2-oxazoline)s suggests them as promising substitutes for poly(ethylene glycol) (PEG) in various applications such as anti-biofouling agents. It was recently reported that poly(2-oxazoline)s themselves have antimicrobial properties as synthetic mimics of host defense peptides. These studies revealed the bioactive properties of poly(2-oxazoline)s as a new class of functional peptide mimics, by mimicking host defense peptides to display potent and selective antimicrobial activities against methicillin-resistant Staphylococcus aureus both in vitro and in vivo, without concerns about antimicrobial resistance. The high structural diversity, facile synthesis, and potent and tunable antimicrobial properties underscore the great potential of poly(2-oxazoline)s as a class of novel antimicrobial agents in dealing with drug-resistant microbial infections and antimicrobial resistance.