一种新的基于SVM相关反馈的图像检索算法

提出了一种结合颜色和形状特征的图像检索方法,针对传统基于内容图像检索不能很好满足用户需求的问题,提出了一种基于支持向量机（SVM）的相关反馈算法来捕捉用户的检索意图。实验结果证明,算法能发挥用户在检索过程中的作用,具有较好检索性能。     

Unifying knowledge iterative dissemination and relational reconstruction network for image–text matching

Image–text matching is a crucial branch in multimedia retrieval which relies on learning inter-modal correspondences. Most existing methods focus on global or local correspondence and fail to explore fine-grained global–local alignment. Moreover, the issue of how to infer more accurate similarity scores remains unresolved. In this study, we propose a novel unifying knowledge iterative dissemination and relational reconstruction (KIDRR) network for image–text matching. Particularly, the knowledge graph iterative dissemination module is designed to iteratively broadcast global semantic knowledge, enabling relevant nodes to be associated, resulting in fine-grained intra-modal correlations and features. Hence, vector-based similarity representations are learned from multiple perspectives to model multi-level alignments comprehensively. The relation graph reconstruction module is further developed to enhance cross-modal correspondences by constructing similarity relation graphs and adaptively reconstructing them. We conducted experiments on the datasets Flickr30K and MSCOCO, which have 31,783 and 123,287 images, respectively. Experiments show that KIDRR achieves improvements of nearly 2.2% and 1.6% relative to Recall@1 on Flicr30K and MSCOCO, respectively, compared to the current state-of-the-art baselines.     

Accelerate adversarial training with loss guided propagation for robust image classification

Adversarial training is effective to train robust image classification models. To improve the robustness, existing approaches often use many propagations to generate adversarial examples, which have high time consumption. In this work, we propose an efficient adversarial training method with loss guided propagation (ATLGP) to accelerate the adversarial training process. ATLGP takes the loss value of generated adversarial examples as guidance to control the number of propagations for each training instance at different training stages, which decreases the computation while keeping the strength of generated adversarial examples. In this way, our method can achieve comparable robustness with less time than traditional training methods. It also has good generalization ability and can be easily combined with other efficient training methods. We conduct comprehensive experiments on CIFAR10 and MNIST, the standard datasets for several benchmarks. The experimental results show that ATLGP reduces 30% to 60% training time compared with other baseline methods while achieving similar robustness against various adversarial attacks. The combination of ATLGP and ATTA (an efficient adversarial training method) achieves superior acceleration potential when robustness meets high requirements. The statistical propagation in different training stages and ablation studies prove the effectiveness of applying loss guided propagation for each training instance. The acceleration technique can more easily extend adversarial training methods to large-scale datasets and more diverse model architectures such as vision transformers.     

匀速直线运动模糊图像的退化模型研究

本文分析并建立了匀速直线运动模糊图像的退化数学模型,研究了点扩展函数的计算方法.结合汽车碰撞试验,对匀速直线运动造成的图像模糊进行了模拟和恢复实验,取得了令人满意的效果.     

Implementing an Image Database for Complex Russian Architectural Objects: The William Brumfield Collection

SUMMARY

With support from a National Endowment for the Humanities Reference Materials Program grant for 2006–2008, faculty and librarians at the University of Washington (UW) are collaborating with William Brumfield (Tulane University) to preserve and catalog the latter's unmatched collection of Russian architectural photographs, create metadata describing the photographs, and make images and text widely accessible as part of an innovative web-based educational and research resource. Building on experience gained through the creation of an experimental pilot database, project staff are adapting emerging standards (METS and CCO) to a custom design that presents images from the Brumfield Collection within their architectural, geographic, and chronological contexts, both as interrelated views of individual structures, and as buildings that share certain structural and stylistic features with other buildings within and beyond the Russian cultural continuum.     

Image Repair on the Donald Trump “Access Hollywood” Video: “Grab Them by the P*ssy”

In October 2016, Donald Trump was in the midst of a hotly contested and sharply divisive presidential campaign. Days before the second presidential debate, The Washington Post posted a video of Trump “having extremely lewd conversation about women in 2005.” This video and the firestorm of criticism it provoked threatened to derail his presidential run. Mr. Trump and his wife Melania Trump offered several messages to repair his damaged image. This article analyzes and evaluates these image repair messages concerning Donald Trump’s “Access Hollywood” video. In such a divisive context, the defense had no hope of repairing Trump’s image with the general public (Trump lost the popular vote by 2.9 million). Even though he lost the popular vote, the business magnate won the Electoral College. We cannot say that the “Access Hollywood” video was solely responsible for Trump’s poor popular vote showing, but it is clear that this defense did not completely dispel the cloud surrounding him.     

Δ-distance: A family of dissimilarity metrics between images represented by multi-level feature vectors

This article presents the Δ-distance, a family of distances between images recursively decomposed into segments and represented by multi-level feature vectors. Such a structure is a quad, a quin or a nona-tree resulting from a fixed and arbitrary image partition or from an image segmentation process. It handles positional information of image features (e.g. color, texture or shape). Δ-distance is the generalized form of dissimilarity measures between multi-level feature vectors. Using different weights on tree nodes and different distances between nodes, distances between trees or visual similarity between images can be computed based on the general definition of Δ. In this article, we present three Δ-based distance families: two families of distances between tree structures, called -distance( for Tree) and -distance ( for Segment), and a family of visual distances between images, called ( for Visual). The -distance visually compares two images using their tree representation and the other two distances compare the tree structures resulting from image segmentation. Moreover, we show how existing distances between multi-level feature vectors appear to be particular cases of the Δ-distance     

基于广义Feistel结构的分组图像加密算法

为了提高加密信息的安全性等要求,提出了一种广义拟feistel结构分组加密算法。首先利用三角混沌映射产生混沌序列并将明文分块,然后将三角混沌映射和灰度扩散机制相结合依次对明文进行替换,最后将三角混沌作为密钥用于广义拟feistel结构分组加密算法对替换后的明文加密。通过对加密后图像相邻像素所对应的共生矩阵特性分析表明,加密图像具有良好的空间随机均匀分布特性。     

《桃夭》的多重隐喻意象

《桃夭》是《诗经》中一首轻快活泼的祝婚诗。全诗三章均以茂盛而美丽的桃树起兴,从修辞学角度讲,属于以物喻人的比喻;从《诗经》的艺术表现手段看则属于兴中兼比,具有明显的隐喻性质。作者选取桃树作诗的整体隐喻意象,以桃花、桃子、桃叶作多重隐喻,层层递进地对新娘婚后生活幸福、早生子女、家业兴旺做了巧妙暗示,同时结合"之子于归,宜其室家"的写实,表达了对新娘幸福美满的婚后生活的热情祝愿。《桃夭》的这种多重隐喻既源于作者对生活的细致观察,又与当时的社会文化背景密切相关。     

中西诗学中的“意象”比较

意象在中国古代文艺理论和西方文论话语中都是十分重要的诗学范畴。出于中国传统文化的“意象”概念与出自西方传统文化的“意象”概念既有相同、相通之处，亦存在明显的差异。人类文化存在着种种共性，在诗歌意象发展过程中，我们可以找到中西诗学意象理论的某些共同规律。同时，由于不同民族的生活习惯、思维方式、自然环境、文化传统等存在着差异性，中西诗学意象理论也存在相异之处，而这种差异在某种意义上说，恰恰构成了一种文化的互补。