基于细粒度视觉特征和知识图谱的视觉故事生成算法

视觉故事生成是图像内容描述衍生的跨模态学习任务,在图文游记自动生成、启蒙教育等领域有较好的应用研究意义。目前主流方法存在对图像细粒度特征描述薄弱、故事文本的图文相关性低、语言不丰富等问题。为此,该文提出了基于细粒度视觉特征和知识图谱的视觉故事生成算法。该算法针对如何对图像内容进行充分挖掘和扩展表示,在视觉和高层语义方面,分别设计实现了图像细粒度视觉特征生成器和图像语义概念词集合生成器两个重要模块。在这两个模块中,细粒度视觉信息通过含有实体关系的场景图结构进行图卷积学习,高层语义信息综合外部知识图谱与相邻图像的语义关联进行扩充丰富,最终实现对图像序列内容较为全面细致的表示。该文算法在目前视觉故事生成领域规模最大的VIST数据集上与主流先进的算法进行了测试。实验结果表明,该文所提算法生成的故事文本,在图文相关性、故事逻辑性、文字多样性等方面,在Distinct-N和TTR等客观指标上均取得较大领先优势,具有良好的应用前景。     

Multimedia ontology matching by using visual and textual modalities

Ontologies have been intensively applied for improving multimedia search and retrieval by providing explicit meaning to visual content. Several multimedia ontologies have been recently proposed as knowledge models suitable for narrowing the well known semantic gap and for enabling the semantic interpretation of images. Since these ontologies have been created in different application contexts, establishing links between them, a task known as ontology matching, promises to fully unlock their potential in support of multimedia search and retrieval. This paper proposes and compares empirically two extensional ontology matching techniques applied to an important semantic image retrieval issue: automatically associating common-sense knowledge to multimedia concepts. First, we extend a previously introduced textual concept matching approach to use both textual and visual representation of images. In addition, a novel matching technique based on a multi-modal graph is proposed. We argue that the textual and visual modalities have to be seen as complementary rather than as exclusive sources of extensional information in order to improve the efficiency of the application of an ontology matching approach in the multimedia domain. An experimental evaluation is included in the paper.     

基于虚拟属性学习的文本-图像行人检索方法

文本-图像行人检索旨在从行人数据库中查找符合特定文本描述的行人图像.近年来受到学术界和工业界的广泛关注.该任务同时面临两个挑战:细粒度检索以及图像与文本之间的异构鸿沟.部分方法提出使用有监督属性学习提取属性相关特征,在细粒度上关联图像和文本.然而属性标签难以获取,导致这类方法在实践中表现不佳.如何在没有属性标注的情况下提取属性相关特征,建立细粒度的跨模态语义关联成为亟待解决的关键问题.为解决这个问题,融合预训练技术提出基于虚拟属性学习的文本-图像行人检索方法,通过无监督属性学习建立细粒度的跨模态语义关联.第一,基于行人属性的不变性和跨模态语义一致性提出语义引导的属性解耦方法,所提方法利用行人的身份标签作为监督信号引导模型解耦属性相关特征.第二,基于属性之间的关联构建语义图提出基于语义推理的特征学习模块,所提模块通过图模型在属性之间交换信息增强特征的跨模态识别能力.在公开的文本-图像行人检索数据集CUHK-PEDES和跨模态检索数据集Flickr30k上与现有方法进行实验对比,实验结果表明了所提方法的有效性.     

部件检测和语义网络的细粒度鞋类图像检索

目的 细粒度图像检索是当前细粒度图像分析和视觉领域的热点问题。以鞋类图像为例,传统方法仅提取其粗粒度特征且缺少关键的语义属性,难以区分部件间的细微差异,不能有效用于细粒度检索。针对鞋类图像检索大多基于简单款式导致检索效率不高的问题,提出一种结合部件检测和语义网络的细粒度鞋类图像检索方法。方法 结合标注后的鞋类图像训练集对输入的待检鞋类图像进行部件检测;基于部件检测后的鞋类图像和定义的语义属性训练语义网络,以提取待检图像和训练图像的特征向量,并采用主成分分析进行降维;通过对鞋类图像训练集中每个候选图像与待检图像间的特征向量进行度量学习,按其匹配度高低顺序输出检索结果。结果 实验在UT-Zap50K数据集上与目前检索效果较好的4种方法进行比较,检索精度提高近6%。同时,与同任务的SHOE-CNN（semantic hierarchy of attribute convolutional neural network）检索方法比较,本文具有更高的检索准确率。结论 针对传统图像特征缺少细微的视觉描述导致鞋类图像检索准确率低的问题,提出一种细粒度鞋类图像检索方法,既提高了鞋类图像检索的精度和准确率,又能较好地满足实际应用需求。     

Cross-modal alignment with graph reasoning for image-text retrieval

Image-text retrieval task has received a lot of attention in the modern research field of artificial intelligence. It still remains challenging since image and text are heterogeneous cross-modal data. The key issue of image-text retrieval is how to learn a common feature space while semantic correspondence between image and text remains. Existing works cannot gain fine cross-modal feature representation because the semantic relation between local features is not effectively utilized and the noise information is not suppressed. In order to address these issues, we propose a Cross-modal Alignment with Graph Reasoning (CAGR) model, in which the refined cross-modal features in the common feature space are learned and then a fine-grained cross-modal alignment method is implemented. Specifically, we introduce a graph reasoning module to explore semantic connection for local elements in each modality and measure their importance by self-attention mechanism. In a multi-step reasoning manner, the visual semantic graph and textual semantic graph can be effectively learned and the refined visual and textual features can be obtained. Finally, to measure the similarity between image and text, a novel alignment approach named cross-modal attentional fine-grained alignment is used to compute similarity score between two sets of features. Our model achieves the competitive performance compared with the state-of-the-art methods on Flickr30K dataset and MS-COCO dataset. Extensive experiments demonstrate the effectiveness of our model.

     

一种基于图注意力的双分支社会关系识别方法

从图像中挖掘人物间的社会关系在刑侦、隐私防护等领域有重要的作用。现有的图建模方法通过创建人际关系图或构建知识图谱来学习人物关系,取得了良好的效果。但基于图卷积神经网络（GCN）的方法一定程度上忽略了不同特征对特定关系的不同程度的重要性。针对上述问题,提出了一种基于图注意力的双分支社会关系识别模型（GAT-DBSR）,第一个分支提取人物区域以及图像全局特征作为节点,核心是通过图注意力网络和门控机制去更新这些节点以学习人物关系的特征表示。第二个分支通过卷积神经网络提取场景特征来增强对人物关系的识别。最终对两个分支的特征进行融合并分类得到所有的社会关系。该模型在PISC数据集的细粒度关系识别任务上的mAP达到了74.4%,相比基线模型提高了1.2%。在PIPA数据集上的关系识别准确率也有一定的提升。实验结果表明了该模型具有更优越的效果。     

Visual and textual fusion for semantically supervised region-based retrieval

This paper presents a unified annotation and retrieval framework, which integrates region annotation with image retrieval for performance reinforcement. To integrate semantic annotation with region-based image retrieval, visual and textual fusion is proposed for both soft matching and Bayesian probabilistic formulations. To address sample insufficiency and sample asymmetry in the annotation classifier training phase, we present a region-level multi-label image annotation scheme based on pair-wise coupling support vector machine (SVM) learning. In the retrieval phase, to achieve semantic-level region matching we present a novel retrieval scheme which differs from former work: the query example uploaded by users is automatically annotated online, and the user can judge its annotation quality. Based on the user’s judgment, two novel schemes are deployed for semantic retrieval: (1) if the user judges the photo to be well annotated, Semantically supervised Integrated Region Matching is adopted, which is a keyword-integrated soft region matching method; (2) If the user judges the photo to be poorly annotated, Keyword Integrated Bayesian Reasoning is adopted, which is a natural integration of a Visual Dictionary in online content-based search. In the relevance feedback phase, we conduct both visual and textual learning to capture the user’s retrieval target. Better annotation and retrieval performance than current methods were reported on both COREL 10,000 and Flickr web image database (25,000 images), which demonstrated the effectiveness of our proposed framework.     

联合姿态先验的人体精确解析双分支网络模型

人体解析旨在将人体图像分割成多个具有细粒度语义的部件区域,进行形成对人体图像的语义理解.然而由于人体姿态的复杂性,现有的人体解析算法容易对人体四肢部件形成误判,且对于小目标区域的分割不够精确.针对上述问题,本文联合人体姿态估计信息,提出了一种人体精确解析的双分支网络模型.该模型首先使用基干网络表征人体图像,将人体姿态估计模型预测到的姿态先验作为基干网络的注意力信息,进而形成人体结构先验驱动的多尺度特征表达,并将提取的特征分别输入至全卷积网络解析分支与检测解析分支.全卷积网络解析分支获得全局分割结果,检测解析分支更关注小尺度目标的检测与分割,融合两个分支的预测信息可获得更为精确的分割结果.实验结果验证了本文算法的有效性,在当前主流的人体解析数据集LIP和ATR上,本文方法的mIoU评测指标分别为52.19%和68.29%,有效提升了解析精度,在人体四肢部件以及小目标部件区域获得了更为准确的分割结果.     

方面级多模态协同注意图卷积情感分析模型

目的 方面级多模态情感分析日益受到关注,其目的是预测多模态数据中所提及的特定方面的情感极性。然而目前的相关方法大都对方面词在上下文建模、模态间细粒度对齐的指向性作用考虑不够,限制了方面级多模态情感分析的性能。为了解决上述问题,提出一个方面级多模态协同注意图卷积情感分析模型（aspect-level multimodal co-attention graph convolutional sentiment analysis model,AMCGC）来同时建模方面指向的模态内上下文语义关联和跨模态的细粒度对齐,以提升情感分析性能。方法 AMCGC为了获得方面导向的模态内的局部语义相关性,利用正交约束的自注意力机制生成各个模态的语义图。然后,通过图卷积获得含有方面词的文本语义图表示和融入方面词的视觉语义图表示,并设计两个不同方向的门控局部跨模态交互机制递进地实现文本语义图表示和视觉语义图表示的细粒度跨模态关联互对齐,从而降低模态间的异构鸿沟。最后,设计方面掩码来选用各模态图表示中方面节点特征作为情感表征,并引入跨模态损失降低异质方面特征的差异。结果 在两个多模态数据集上与9种方法进行对比,在Twitter-2015数据集中,相比于性能第2的模型,准确率提高了1.76%;在Twitter-2017数据集中,相比于性能第2的模型,准确率提高了1.19%。在消融实验部分则从正交约束、跨模态损失、交叉协同多模态融合分别进行评估,验证了AMCGC模型各部分的合理性。结论 本文提出的AMCGC模型能更好地捕捉模态内的局部语义相关性和模态之间的细粒度对齐,提升方面级多模态情感分析的准确性。     

基于哈希特征的大规模图像快速匹配算法

针对基于图像进行三维重建技术在使用大规模图像集合进行重建时,需要对图像集合中图像进行两两匹配耗时问题,提出了基于哈希技术对图像构建全局哈希特征的方法,通过过滤掉无效的图像关系对来减少计算时间,极大地提高了大规模图像集合三维重建的匹配计算效率。提出的大规模图像快速哈希匹配算法包括构建图像哈希特征、构建初始匹配图、挑选候选匹配对、哈希匹配几个步骤。实验结果表明该方法能显著地提高三维重建中图像匹配的速度。     

结合细粒度特征与深度卷积网络的手绘图检索

目的 传统的手绘图像检索方法主要集中在检索相同类别的图像，忽略了手绘图像的细粒度特征。对此，提出了一种新的结合细粒度特征与深度卷积网络的手绘图像检索方法，既注重通过深度跨域实现整体匹配，也实现细粒度细节匹配。方法 首先构建多通道混合卷积神经网络，对手绘图像和自然图像分别进行不同的处理；其次通过在网络中加入注意力模型来获取细粒度特征；最后将粗细特征融合，进行相似性度量，得到检索结果。结果 在不同的数据库上进行实验，与传统的尺度不变特征（SIFT）、方向梯度直方图（HOG）和深度手绘模型Deep SaN（sketch-a-net）、Deep 3DS（sketch）、Deep TSN（triplet sketch net）等5种基准方法进行比较，选取了Top-1和Top-10，在鞋子数据集上，本文方法Top-1正确率提升了12%，在椅子数据集上，本文方法Top-1正确率提升了11%，Top-10提升了3%，与传统的手绘检索方法相比，本文方法得到了更高的准确率。在实验中，本文方法通过手绘图像能在第1幅检索出绝大多数的目标图像，达到了实例级别手绘检索的目的。结论 提出了一种新的手绘图像检索方法，为手绘图像和自然图像的跨域检索提供了一种新思路，进行实例级别的手绘检索，与原有的方法相比，检索精度得到明显提升，证明了本文方法的可行性。     

Structure-sensitive semantic matching for aggregate question answering over knowledge base

Aggregate question answering essentially returns answers for given questions by obtaining query graphs with unique dependencies between values and corresponding objects. Word order dependency, as the key to uniquely identify dependency of the query graph, reflects the dependencies between the words in the question. However, due to the semantic gap caused by the expression difference between questions encoded with word vectors and query graphs represented with logical formal elements, it is not trivial to match the correct query graph for the question. Most existing approaches design more expressive query graphs for complex questions and rank them just by directly calculating their similarities, ignoring the semantic gap between them. In this paper, we propose a novel Structure-sensitive Semantic Matching(SSM) approach that learns aligned representations of dependencies in questions and query graphs to eliminate their gap. First, we propose a cross-structure matching module to bridge the gap between two modalities(i.e., textual question and query graph). Then, we propose an entropy-based gated AQG filter to remove the structural noise caused by the uncertainty of dependencies. Finally, we present a two-channel query graph representation that fuses the semantics of abstract structure and grounding content of the query graph explicitly. Experimental results show that SSM could learn aligned representations of questions and query graphs to eliminate the gaps between their dependencies, and improves up to 12% (F1 score) on aggregation questions of two benchmark datasets.     

Part and Attribute Discovery from Relative Annotations

Part and attribute based representations are widely used to support high-level search and retrieval applications. However, learning computer vision models for automatically extracting these from images requires significant effort in the form of part and attribute labels and annotations. We propose an annotation framework based on comparisons between pairs of instances within a set, which aims to reduce the overhead in manually specifying the set of part and attribute labels. Our comparisons are based on intuitive properties such as correspondences and differences, which are applicable to a wide range of categories. Moreover, they require few category specific instructions and lead to simple annotation interfaces compared to traditional approaches. On a number of visual categories we show that our framework can use noisy annotations collected via “crowdsourcing” to discover semantic parts useful for detection and parsing, as well as attributes suitable for fine-grained recognition.     

无监督的中文商品属性结构化方法

从非结构化商品描述文本中抽取结构化属性信息,对于电子商务实现商品的对比与推荐及用户需求预测等功能具有重要意义.现有结构化方法大多采用监督或半监督的分类方法抽取属性值与属性名,通过文法分析器分析属性值与属性名之间的文法依存关系,并根据关联规则实现属性值与属性名的匹配.这些方法存在以下不足：（1）需要人工标记部分属性值、属性名及它们之间的对应关系;（2）属性值-属性名匹配的准确度受到语言习惯、句意逻辑、语料库及属性名候选集质量的严重制约.提出了一种无监督的中文商品属性结构化方法.该方法借助搜索引擎,基于小概率事件原理分析文法关系来抽取属性值与属性名.同时,提出相对不选取条件概率场,并使用Page Rank算法来计算属性值与属性名的配对概率.该方法无需人工标记的开销,且无论商品描述中是否显式地包含相应的属性名,该方法都能自动抽取到属性值并匹配相应的属性名.使用百度搜索引擎上的真实语料,针对4类商品的中文描述进行了实验.实验结果验证了对于候选属性名的自动生成,所提出的基于搜索引擎搜索属性值,并在包含属性值的搜索结果中抽取一般名词的候选属性名生成方法与只在描述句中抽取一般名词的候选属性名生成方法相比,查全率提高了20%以上;对于非量化类属性,所提出的基于相对不选取条件概率场的属性值-属性名匹配方法与基于依存关联的方法相比,Rank-1的准确率提高了30%以上,平均MRR提高了0.3以上.