Multiple level visual semantic fusion method for image re-ranking

Mid-level semantic attributes have obtained some success in image retrieval and re-ranking. However, due to the semantic gap between the low-level feature and intermediate semantic concept, information loss is considerable in the process of converting the low-level feature to semantic concept. To tackle this problem, we tried to bridge the semantic gap by looking for the complementary of different mid-level features. In this paper, a framework is proposed to improve image re-ranking by fusing multiple mid-level features together. The framework contains three mid-level features (DCNN-ImageNet attributes, Fisher vector, sparse coding spatial pyramid matching) and a semi-supervised multigraph-based model that combines these features together. In addition, our framework can be easily extended to utilize arbitrary number of features for image re-ranking. The experiments are conducted on the a-Pascal dataset, and our approach that fuses different features together is able to boost performance of image re-ranking efficiently.     

Object Recognition as Many-to-Many Feature Matching

Object recognition can be formulated as matching image features to model features. When recognition is exemplar-based, feature correspondence is one-to-one. However, segmentation errors, articulation, scale difference, and within-class deformation can yield image and model features which don’t match one-to-one but rather many-to-many. Adopting a graph-based representation of a set of features, we present a matching algorithm that establishes many-to-many correspondences between the nodes of two noisy, vertex-labeled weighted graphs. Our approach reduces the problem of many-to-many matching of weighted graphs to that of many-to-many matching of weighted point sets in a normed vector space. This is accomplished by embedding the initial weighted graphs into a normed vector space with low distortion using a novel embedding technique based on a spherical encoding of graph structure. Many-to-many vector correspondences established by the Earth Mover’s Distance framework are mapped back into many-to-many correspondences between graph nodes. Empirical evaluation of the algorithm on an extensive set of recognition trials, including a comparison with two competing graph matching approaches, demonstrates both the robustness and efficacy of the overall approach.     

Loop Closure Detection via Locality Preserving Matching With Global Consensus

A critical component of visual simultaneous localization and mapping is loop closure detection (LCD), an operation judging whether a robot has come to a pre-visited area. Concretely, given a query image (i.e., the latest view observed by the robot), it proceeds by first exploring images with similar semantic information, followed by solving the relative relationship between candidate pairs in the 3D space. In this work, a novel appearance-based LCD system is proposed. Specifically, candidate frame selection is conducted via the combination of Super-features and aggregated selective match kernel (ASMK). We incorporate an incremental strategy into the vanilla ASMK to make it applied in the LCD task. It is demonstrated that this setting is memory-wise efficient and can achieve remarkable performance. To dig up consistent geometry between image pairs during loop closure verification, we propose a simple yet surprisingly effective feature matching algorithm, termed locality preserving matching with global consensus (LPM-GC). The major objective of LPM-GC is to retain the local neighborhood information of true feature correspondences between candidate pairs, where a global constraint is further designed to effectively remove false correspondences in challenging sceneries, e.g., containing numerous repetitive structures. Meanwhile, we derive a closed-form solution that enables our approach to provide reliable correspondences within only a few milliseconds. The performance of the proposed approach has been experimentally evaluated on ten publicly available and challenging datasets. Results show that our method can achieve better performance over the state-of-the-art in both feature matching and LCD tasks. We have released our code of LPM-GC at https://github.com/jiayi-ma/LPM-GC.      

Robust Structure‐Based Shape Correspondence

We present a robust method to find region‐level correspondences between shapes, which are invariant to changes in geometry and applicable across multiple shape representations. We generate simplified shape graphs by jointly decomposing the shapes, and devise an adapted graph‐matching technique, from which we infer correspondences between shape regions. The simplified shape graphs are designed to primarily capture the overall structure of the shapes, without reflecting precise information about the geometry of each region, which enables us to find correspondences between shapes that might have significant geometric differences. Moreover, due to the special care we take to ensure the robustness of each part of our pipeline, our method can find correspondences between shapes with different representations, such as triangular meshes and point clouds. We demonstrate that the region‐wise matching that we obtain can be used to find correspondences between feature points, reveal the intrinsic self‐similarities of each shape and even construct point‐to‐point maps across shapes. Our method is both time and space efficient, leading to a pipeline that is significantly faster than comparable approaches. We demonstrate the performance of our approach through an extensive quantitative and qualitative evaluation on several benchmarks where we achieve comparable or superior performance to existing methods.     

哈希编码结合空间金字塔的图像分类

目的 稀疏编码是当前广泛使用的一种图像表示方法,针对稀疏编码及其改进算法计算过程复杂、费时等问题,提出一种哈希编码结合空间金字塔的图像分类算法。方法 首先,提取图像的局部特征点,构成局部特征点描述集。其次,学习自编码哈希函数,将局部特征点表示为二进制哈希编码。然后,在二进制哈希编码的基础上进行K均值聚类生成二进制视觉词典。最后,结合空间金字塔模型,将图像表示为空间金字塔直方图向量,并应用于图像分类。结果 在常用的Caltech-101和Scene-15数据集上进行实验验证,并和目前与稀疏编码相关的算法进行实验对比。与稀疏编码相关的算法相比,本文算法词典学习时间缩短了50%,在线编码速度提高了1.3～12.4倍,分类正确率提高了1%～5%。结论 提出了一种哈希编码结合空间金字塔的图像分类算法,利用哈希编码代替稀疏编码对局部特征点进行编码,并结合空间金字塔模型用于图像分类。实验结果表明,本文算法词典学习时间更短、编码速度更快,适用于在线词典学习和应用。     

Speaker recognition using pyramid match kernel based support vector machines

Gaussian mixture model (GMM) based approaches have been commonly used for speaker recognition tasks. Methods for estimation of parameters of GMMs include the expectation-maximization method which is a non-discriminative learning based method. Discriminative classifier based approaches to speaker recognition include support vector machine (SVM) based classifiers using dynamic kernels such as generalized linear discriminant sequence kernel, probabilistic sequence kernel, GMM supervector kernel, GMM-UBM mean interval kernel (GUMI) and intermediate matching kernel. Recently, the pyramid match kernel (PMK) using grids in the feature space as histogram bins and vocabulary-guided PMK (VGPMK) using clusters in the feature space as histogram bins have been proposed for recognition of objects in an image represented as a set of local feature vectors. In PMK, a set of feature vectors is mapped onto a multi-resolution histogram pyramid. The kernel is computed between a pair of examples by comparing the pyramids using a weighted histogram intersection function at each level of pyramid. We propose to use the PMK-based SVM classifier for speaker identification and verification from the speech signal of an utterance represented as a set of local feature vectors. The main issue in building the PMK-based SVM classifier is construction of a pyramid of histograms. We first propose to form hard clusters, using k-means clustering method, with increasing number of clusters at different levels of pyramid to design the codebook-based PMK (CBPMK). Then we propose the GMM-based PMK (GMMPMK) that uses soft clustering. We compare the performance of the GMM-based approaches, and the PMK and other dynamic kernel SVM-based approaches to speaker identification and verification. The 2002 and 2003 NIST speaker recognition corpora are used in evaluation of different approaches to speaker identification and verification. Results of our studies show that the dynamic kernel SVM-based approaches give a significantly better performance than the state-of-the-art GMM-based approaches. For speaker recognition task, the GMMPMK-based SVM gives a performance that is better than that of SVMs using many other dynamic kernels and comparable to that of SVMs using state-of-the-art dynamic kernel, GUMI kernel. The storage requirements of the GMMPMK-based SVMs are less than that of SVMs using any other dynamic kernel.     

基于多尺度特征映射匹配的图像表示方法

在卷积神经网络模型中,空间金字塔池化方法将空间信息融入到深度特征的生成过程中,最终生成的图像表示可以有效地用于提高图像检索性能,但是此方法会导致生成的图像表示中不同维度之间描述的信息存在重复且相同维度描述的图像内容不匹配。为此提出了一种基于多尺度特征映射匹配（multi-scale feature map matching,MFMM）的图像表示方法,此方法首先利用深度特征的方差与协方差矩阵提出了一种特征映射选择算法,用于增强图像表示中不同维度特征的独立性。其次,依据相同通道特征映射中高响应值位置有较高匹配性的特点,结合激活映射中最大响应位置的深度特征提出了一种优化的特征映射中心点选择方法。最后,按照不同的中心点通过多尺度窗口采样的方式,从特征映射中提取出带有空间信息的深度特征用于表示图像内容。实验结果表明,提出的方法在图像检索任务中能够取得良好的效果。     

Global Correspondence Optimization for Non‐Rigid Registration of Depth Scans

We present a registration algorithm for pairs of deforming and partial range scans that addresses the challenges of non‐rigid registration within a single non‐linear optimization. Our algorithm simultaneously solves for correspondences between points on source and target scans, confidence weights that measure the reliability of each correspondence and identify non‐overlapping areas, and a warping field that brings the source scan into alignment with the target geometry. The optimization maximizes the region of overlap and the spatial coherence of the deformation while minimizing registration error. All optimization parameters are chosen automatically; hand‐tuning is not necessary. Our method is not restricted to part‐in‐whole matching, but addresses the general problem of partial matching, and requires no explicit prior correspondences or feature points. We evaluate the performance and robustness of our method using scan data acquired by a structured light scanner and compare our method with existing non‐rigid registration algorithms.     

Unsupervised Learning for Graph Matching

Graph matching is an essential problem in computer vision that has been successfully applied to 2D and 3D feature matching and object recognition. Despite its importance, little has been published on learning the parameters that control graph matching, even though learning has been shown to be vital for improving the matching rate. In this paper we show how to perform parameter learning in an unsupervised fashion, that is when no correct correspondences between graphs are given during training. Our experiments reveal that unsupervised learning compares favorably to the supervised case, both in terms of efficiency and quality, while avoiding the tedious manual labeling of ground truth correspondences. We verify experimentally that our learning method can improve the performance of several state-of-the art graph matching algorithms. We also show that a similar method can be successfully applied to parameter learning for graphical models and demonstrate its effectiveness empirically.     

基于极线校正的最小二乘立体匹配

本文减少了最小二乘转换参数,通过三个几何转换参数和两个辐射转换参数建立对应关系,采用经极线校正的立体像对,使对应点的搜索在相同扫描行上进行,减小了搜索空间,提高了匹配速度,且把匹配方法嵌入到多尺度空间中以提高匹配速度,通过视差后处理进一步提高匹配精度。采用自适应窗口技术解决由于存在矩阵不可逆情况而导致大量不可匹配点和在地形平坦、灰度变化不明显的区域不匹配或误匹配率高的缺点。试验结果表明了,本算法精度高,匹配率高的优点,有相当的使用价值。     

Fast verification via statistical geometric for mobile visual search

In this paper, an efficient geometric statistics method is proposed to obtain the geometric information of the object, which can achieve fast visual re-ranking along with the localization of target-of-interest. Given an input pair of images, first we get a set of interest-point correspondences, and enumerate all potential pairs in each image, upon which we calculate the statistics of the corresponding pairs to yield the geometric similarity score. We use a location geometric similarity scoring method that is invariant to rotation, scale, and translation, and can be easily incorporated in mobile visual search and augmented reality systems. Then fitting the statistics of geometric similarity scores into a Gaussian distribution that is used as a priori to determine the matching. The performance of our geometric scoring scheme is compared to the conventional geometric scoring schemes using orientation and scale. It is shown that our proposed statistically geometric method can generate fast geometric re-ranking. Meanwhile, we can accurately locate the target of search interest regardless of variations caused by occlusion and perspective changes.     

A new point matching algorithm for non-rigid registration

Feature-based methods for non-rigid registration frequently encounter the correspondence problem. Regardless of whether points, lines, curves or surface parameterizations are used, feature-based non-rigid matching requires us to automatically solve for correspondences between two sets of features. In addition, there could be many features in either set that have no counterparts in the other. This outlier rejection problem further complicates an already difficult correspondence problem. We formulate feature-based non-rigid registration as a non-rigid point matching problem. After a careful review of the problem and an in-depth examination of two types of methods previously designed for rigid robust point matching (RPM), we propose a new general framework for non-rigid point matching. We consider it a general framework because it does not depend on any particular form of spatial mapping. We have also developed an algorithm—the TPS–RPM algorithm—with the thin-plate spline (TPS) as the parameterization of the non-rigid spatial mapping and the softassign for the correspondence. The performance of the TPS–RPM algorithm is demonstrated and validated in a series of carefully designed synthetic experiments. In each of these experiments, an empirical comparison with the popular iterated closest point (ICP) algorithm is also provided. Finally, we apply the algorithm to the problem of non-rigid registration of cortical anatomical structures which is required in brain mapping. While these results are somewhat preliminary, they clearly demonstrate the applicability of our approach to real world tasks involving feature-based non-rigid registration.