Manifold Ranking Research Articles

Image retrieval framework based on texton uniform descriptor and modified manifold ranking

Image representation and ranking are crucial parts in image retrieval. These two steps are independently constructed in most retrieval models, but the compatibility between descriptors and ranking algorithms play an important role. Inspired by human vision perception and manifold learning, we propose a novel image retrieval framework in this paper. We first propose an image representation called texton uniform descriptor, and then illustrate the preservation of the intrinsic manifold structure through visualizing the distribution of image representations on the two-dimensional manifold. This characteristic provides the foundation for subsequent manifold-based ranking. To further improve the efficiency in image retrieval, we propose modified manifold ranking (MMR) which aims at selecting small-scale images randomly as landmarks to propagate adjacent similarity among images iteratively. The extensive experiments in four public datasets demonstrate that our framework has better performance than other state-of-the-art methods in image retrieval.

Robust saliency detection via corner information and an energy function

In this study, the authors propose a distinctive bottom‐up visual saliency detection algorithm based on a new background prior and a new reinforcement. Inspired by genetic algorithm, the final map is obtained with three steps. First of all, the authors construct a background‐based saliency map by manifold ranking via superior image corners selected by convex‐hull as background prior, which is different from most of the existing background prior‐based methods treated all image boundaries as background. Then, a better result is obtained by ranking the relevance of the image elements with foreground seeds extracted from the preliminary saliency map. Furthermore, a novel optimisation framework is introduced with the intention of refining the map, which integrates an energy function with a guided filter. Experimental results on three public datasets indicate that the proposed method performs favourably against the state‐of‐the‐art algorithms.

Moving object detection via robust background modeling with recurring patterns voting

This paper proposes a simple yet effective background modeling method based on recurring patterns voting for moving object detection under challenging scenes. Our method performs the following two steps. First, we employ Gaussian Mixture Model (GMM) to generate the initial background probability map for each frame, in which the value of each pixel represents its probability belonging to the background. Second, we perform recurring patterns voting by employing the graph-based manifold ranking algorithm on the spatially constrained graph to refine the probability map. This prior bases on the observation that the same patterns tend to recur frequently in same semantic regions (background or foreground). We verify it in our problem by calculating the background-background, foreground-foreground and background-foreground densities on some video sequences with ground truths. Experimental results on public video sequences suggest that the proposed method significantly outperforms other moving object detection methods.

Saliency detection by hierarchically integrating compactness, contrast and boundary connectivity

Saliency detection is one of the most challenging problems in computer vision and has extensive applications in many fields. In this work, instead of simply defining the compactness and contrast, we design novel versions of these two cues based on manifold ranking, and then propose a saliency detection model by integrating the newly modified compactness and contrast with boundary connectivity. Since various scales salient detections highlight different parts of the objects, to further improve the performance, we perform the model hierarchically on four different scales and then fuse the results to obtain the final saliency map. Experiments on four benchmark datasets demonstrate the effectiveness of the proposed method. The method can further improve the accuracy of saliency detection than other 15 state-of-the-art methods on MSRA10k, ASD, DUT-OMRON and ECSSD.

Learning Dual Multi-Scale Manifold Ranking for Semantic Segmentation of High-Resolution Images

Semantic image segmentation has recently witnessed considerable progress by training deep convolutional neural networks (CNNs). The core issue of this technique is the limited capacity of CNNs to depict visual objects. Existing approaches tend to utilize approximate inference in a discrete domain or additional aides and do not have a global optimum guarantee. We propose the use of the multi-label manifold ranking (MR) method in solving the linear objective energy function in a continuous domain to delineate visual objects and solve these problems. We present a novel embedded single stream optimization method based on the MR model to avoid approximations without sacrificing expressive power. In addition, we propose a novel network, which we refer to as dual multi-scale manifold ranking (DMSMR) network, that combines the dilated, multi-scale strategies with the single stream MR optimization method in the deep learning architecture to further improve the performance. Experiments on high resolution images, including close-range and remote sensing datasets, demonstrate that the proposed approach can achieve competitive accuracy without additional aides in an end-to-end manner.

A novel classification method for aid decision of traditional Chinese patent medicines for stroke treatment.

Traditional Chinese patent medicines are widely used to treat stroke because it has good efficacy in the clinical environment. However, because of the lack of knowledge on traditional Chinese patent medicines, many Western physicians, who are accountable for the majority of clinical prescriptions for such medicine, are confused with the use of traditional Chinese patent medicines. Therefore, the aid-decision method is critical and necessary to help Western physicians rationally use traditional Chinese patent medicines. In this paper, Manifold Ranking is employed to develop the aid-decision model of traditional Chinese patent medicines for stroke treatment. First, 115 stroke patients from three hospitals are recruited in the cross-sectional survey. Simultaneously, traditional Chinese physicians determine the traditional Chinese patent medicines appropriate for each patient. Second, particular indicators are explored to characterize the population feature of traditional Chinese patent medicines for stroke treatment. Moreover, these particular indicators can be easily obtained byWestern physicians and are feasible for widespread clinical application in the future. Third, the aid-decision model of traditional Chinese patent medicines for stroke treatment is constructed based on Manifold Ranking. Experimental results reveal that traditional Chinese patent medicines can be differentiated. Moreover, the proposed model can obtain high accuracy of aid decision.

Salient object detection via color and texture cues

In this paper, we present a new bottom-up salient object detection approach by constructing two graphs using color and texture features within the manifold ranking framework. First, we calculate the saliency of boundary patches and exclude the ones with high saliency which might be a part of saliency object. Second, we adopt a two-stage scheme for salient detection via affinity propagation clustering and graph-based manifold ranking. The background-based saliency detection aims to obtain the salient object regions as much as possible. In the foreground-based saliency detection, a similar computation is processed as that in the former step and yet slightly different. Instead of simultaneously using all the extracted boundary patches or foreground patches as queries, we compute saliency by using the patches in each cluster in turn and integrating them. At last, the final saliency map is generated by linearly combining two saliency maps respectively exploring color and texture cues. Both qualitative and quantitative evaluations on three publicly available datasets demonstrate the robustness and efficiency of our proposed approach against 21 state-of-the-art methods.

Salient object detection with high‐level prior based on Bayesian fusion

Most of approaches to salient object detection focused on two‐dimensional images, while rare attention was attached to the light field which can provide exclusive visual information for salient object detection and other computer vision applications. An effective algorithm of salient object detection is proposed for light field data. First, boundary connectivity is calculated on all‐focus image. Then, background probability based on boundary connectivity is achieved by computing geodesic distance. Second, the authors rank the similarity of the superpixels of both all‐focus image and depth map via graph‐based manifold ranking to carry out two initial saliency maps. Third, weighted by background probability, the two initial saliency maps are fused to produce final saliency results, integrated by objectness cue. The authors also exploit how to integrate effectively objectness with other visual features, and compare two fusion strategies: linear fusion and Bayesian integration. Experiments show that light field features are helpful for saliency detection, and Bayesian integration framework is the better choice than linear fusion method. Meanwhile, the way how to combine multiple features is crucial. The proposed algorithm handles challenging natural scenarios such as cluttered background, similar foreground and background, and so on, and produces visual favourable results in comparison with the eight state‐of‐the‐art methods.

Salient object detection via compactness and objectness cues

Existing saliency detection algorithms are mainly patch-based. In this paper, we propose a simple but effective approach to detect salient objects by exploring both patch-level and object-level cues. First, we obtain the objectness saliency map with objectness algorithm to find potential object candidates without need of category information. Second, the compactness map is generated by measuring color spatial distribution, and then it is refined by eliminating regions connecting to the selected boundary. Finally, to enforce the consistency among salient regions, we adopt graph-based manifold ranking algorithm by constructing two graphs each using a regional property descriptor. Both qualitative and quantitative evaluations on four publicly available datasets demonstrate the robustness and efficiency of our proposed approach against 23 state-of-the-art methods in terms of six performance criterions.

Rapid three-dimensional scene modeling by sketch retrieval and auto-arrangement

The existing three-dimensional (3D) object layout methods are focused mainly on indoor scenes and they are limited for outdoor applications. In this study, we propose a data-driven method for outdoor scene modeling by using fast retrieval and automatic optimization layout techniques. Unlike the current methods, we first employ an improved manifold ranking algorithm in the sketch-based 3D model retrieval stage, which achieves higher accuracy. Next, according to the particular properties of outdoor architectures, specialized constraints are then proposed to define an energy function, which meets both the functional and aesthetic requirements. Finally, we cast the auto-arrangement as a combinatorial optimization problem, which we solve using an optimization algorithm. In contrast to the earlier version of this method, which was presented at Cyberworlds 2016, this extended version combines simulated annealing and particle swarm optimization algorithms, which have the advantages of rapid convergence and avoiding becoming trapped by local minima. Our experimental results demonstrate that the proposed method is more intuitive and effective for modeling 3D scenes, and can be employed in the actual development of game scenes.

Image Alignment-Based Multi-Region Matching for Object-Level Tampering Detection

Tampering detection methods based on image hashing have been widely studied with continuous advancements. However, most existing models cannot generate object-level tampering localization results, because the forensic hashes attached to the image lack contour information. In this paper, we present a novel tampering detection model that can generate an accurate, object-level tampering localization result. First, an adaptive image segmentation method is proposed to segment the image into closed regions based on strong edges. Then, the color and position features of the closed regions are extracted as a forensic hash. Furthermore, a geometric invariant tampering localization model named image alignment-based multi-region matching (IAMRM) is proposed to establish the region correspondence between the received and forensic images by exploiting their intrinsic structure information. The model estimates the parameters of geometric transformations via a robust image alignment method based on triangle similarity; in addition, it matches multiple regions simultaneously by utilizing manifold ranking based on different graph structures and features. Experimental results demonstrate that the proposed IAMRM is a promising method for object-level tampering detection compared with the state-of-the-art methods.

Manifold-ranking embedded order preserving hashing for image semantic retrieval

Due to the storage and computational efficiency of hashing technology, it has proven a valuable tool for large scale similarity search. In many cases, the large scale data in real-world lie near some (unknown) low-dimensional and non-linear manifold. Moreover, Manifold Ranking approach can preserve the global topological structure of the data set more effectively than Euclidean Distance-based Ranking approach, which fails to preserve the semantic relevance degree. However, most existing hashing methods ignore the global topological structure of the data set. The key issue is how to incorporate the global topological structure of data set into learning effective hashing function. In this paper, we propose a novel unsupervised hashing approach, namely Manifold-Ranking Embedded Order Preserving Hashing (MREOPH). A manifold ranking loss is introduced to solve the issue of global topological structure preserving. An order preserving loss is introduced to ensure the consistency between manifold ranking and hamming ranking. A hypercubic quantization loss is introduced to learn discrete binary codes. The information theoretic regularization term is taken into consideration for preserving desirable properties of hash codes. Finally, we integrate them in a joint optimization framework for minimizing the information loss in each processing. Experimental results on three datasets for semantic search clearly demonstrate the effectiveness of the proposed method.

Heterogeneous Manifold Ranking for Image Retrieval

Graph-based ranking models, such as manifold ranking (MR), have been widely used in various image retrieval applications. To further improve such models, a current trend is to fuse the ranking results from multiple feature sets. Most of existing methods mainly concentrate on fusing the homogeneous feature sets derived from a single information channel, like the multiple modalities of image visual content, but little is known in fusing such heterogeneous feature sets derived from multiple information channels as the click-through data associated with images and their visual content. The primary challenge is how to effectively exploit the complementary properties of the heterogeneous feature sets. Another tough issue is the low-quality nature of the click-through data, which makes the exploration of such complementary properties more difficult. In this paper, we propose a heterogeneous MR (HMR) model, in which a couple of graphs built on the click and visual feature sets are fused to simultaneously encode the image ranking results. Specifically, our HMR model applies different solutions to fuse the heterogeneous feature sets in terms of whether the relevance feedback mechanism is available or not. In addition, we develop a click refinement technique to address the noiseness and sparseness problems inherent in the click-through data. Concretely, it prunes the inaccurate clicks from the click-through data using a neighbor voting strategy, and then enriches the pruned data with novel yet accurate clicks based on a novel collaborative filtering (CF) approach, which is devised by integrating the merits of three popularly used CF methods, thus called Tri-CF algorithm. Extensive experiments on the tasks of click refinement and image retrieval demonstrate the superior performance of the proposed algorithms over several representative methods, especially when the click-through data is highly noisy and sparse.

A manifold ranking based method using hybrid features for crime scene shoeprint retrieval

Shoeprints are frequently acquired at crime scenes and can provide vital clues for investigation. How to retrieve the most similar shoeprints available in the dataset to the highly degraded query crime scene shoeprint is a challenging work. Some existing shoeprint retrieval algorithms cannot work well with highly degraded shoeprint image on a large scale dataset, and the results are not well correlated with the forensic experts. This study proposes a manifold ranking based method using hybrid features of region and appearance to improve the retrieval performance. Manifold ranking method is introduced to narrow the well-known gap between visual features and semantic concepts. We define the ranking cost function from three perspectives: (i) the feature similarity between the query and the dataset images, (ii) the relationship between every two shoeprints in the dataset, (iii) the assigned opinion scores for multiple shoeprints left in one crime scene by the forensic expert. Experiments on the real crime scene datasets show that the cumulative match score of the proposed method is more than 93.5 % on the top 2 % of the dataset composed of 10096 crime scene shoeprints.

Note: A manifold ranking based saliency detection method for camera.

Research focused on salient object region in natural scenes has attracted a lot in computer vision and has widely been used in many applications like object detection and segmentation. However, an accurate focusing on the salient region, while taking photographs of the real-world scenery, is still a challenging task. In order to deal with the problem, this paper presents a novel approach based on human visual system, which works better with the usage of both background prior and compactness prior. In the proposed method, we eliminate the unsuitable boundary with a fixed threshold to optimize the image boundary selection which can provide more precise estimations. Then, the object detection, which is optimized with compactness prior, is obtained by ranking with background queries. Salient objects are generally grouped together into connected areas that have compact spatial distributions. The experimental results on three public datasets demonstrate that the precision and robustness of the proposed algorithm have been improved obviously.

Automated layer segmentation of macular OCT images via graph-based SLIC superpixels and manifold ranking approach

Using the graph-based a simple linear iterative clustering (SLIC) superpixels and manifold ranking technology, a novel automated intra-retinal layer segmentation method is proposed in this paper. Eleven boundaries of ten retinal layers in optical coherence tomography (OCT) images are exactly, fast and reliably quantified. Instead of considering the intensity or gradient features of the single-pixel in most existing segmentation methods, the proposed method focuses on the superpixels and the connected components-based image cues. The image is represented as some weighted graphs with superpixels or connected components as nodes. Each node is ranked with the gradient and spatial distance cues via graph-based Dijkstra’s method or manifold ranking. So that it can effectively overcome speckle noise, organic texture and blood vessel artifacts issues. Segmentation is carried out in a three-stage scheme to extract eleven boundaries efficiently. The segmentation algorithm is validated on 2D and 3D OCT images in three databases, and is compared with the manual tracings of two independent observers. It demonstrates promising results in term of the mean unsigned boundaries errors, the mean signed boundaries errors, and layers thickness errors.

Salient object detection in complex scenes via D-S evidence theory based region classification

In complex scenes, multiple objects are often concealed in cluttered backgrounds. Their saliency is difficult to be detected by using conventional methods, mainly because single color contrast can not shoulder the mission of saliency measure; other image features should be involved in saliency detection to obtain more accurate results. Using Dempster-Shafer (D-S) evidence theory based region classification, a novel method is presented in this paper. In the proposed framework, depth feature information extracted from a coarse map is employed to generate initial feature evidences which indicate the probabilities of regions belonging to foreground or background. Based on the D-S evidence theory, both uncertainty and imprecision are modeled, and the conflicts between different feature evidences are properly resolved. Moreover, the method can automatically determine the mass functions of the two-stage evidence fusion for region classification. According to the classification result and region relevance, a more precise saliency map can then be generated by manifold ranking. To further improve the detection results, a guided filter is utilized to optimize the saliency map. Both qualitative and quantitative evaluations on three publicly challenging benchmark datasets demonstrate that the proposed method outperforms the contrast state-of-the-art methods, especially for detection in complex scenes.

A SYSTEMATIC SEARCH FOR THE SPECTRA WITH FEATURES OF CRYSTALLINE SILICATES IN THE SPITZER IRS ENHANCED PRODUCTS

ABSTRACT The crystalline silicate features are mainly reflected in infrared bands. The Spitzer Infrared Spectrograph (IRS) collected numerous spectra of various objects and provided a big database to investigate crystalline silicates in a wide range of astronomical environments. We apply the manifold ranking algorithm to perform a systematic search for the spectra with crystalline silicate features in the Spitzer IRS Enhanced Products available. In total, 868 spectra of 790 sources are found to show the features of crystalline silicates. These objects are cross-matched with the SIMBAD database as well as with the Large Sky Area Multi-object Fiber Spectroscopic Telescope (LAMOST)/DR2. The average spectrum of young stellar objects shows a variety of features dominated either by forsterite or enstatite or neither, while the average spectrum of evolved objects consistently present dominant features of forsterite in AGB, OH/IR, post-AGB, and planetary nebulae. They are identified optically as early-type stars, evolved stars, galaxies and so on. In addition, the strength of spectral features in typical silicate complexes is calculated. The results are available through CDS for the astronomical community to further study crystalline silicates.

Discriminative saliency propagation with sink points

Salient object detection is still very challenging especially in images with complex or cluttered background. In this paper, we present an efficient and discriminative framework to address it. In specially, a discriminative similarity metric is first proposed by measuring the chi-square distance in a new constructed feature space. Then, we apply it to calculate a background based coarse saliency map by introducing distribution prior to remove foreground noises in the image boundaries. Based on manifold ranking, a robust saliency propagation mechanism is further developed to highlight salient object and simultaneously suppress background region by setting appropriate sink points. Finally, several simple refinement techniques are utilized to generate pixel-wise and smooth saliency maps. Extensive experimental results show the superior performance of the proposed method in terms of different evaluation metrics. In addition, the proposed framework can be also applied to the existing saliency propagation methods for significant performance boosting. We also believe that it is a good choice for subsequent applications based on the achieved high performance and acceptable computational overhead.

Fuzzy evolutionary cellular learning automata model for text summarization

Abstract Text summarization is the automatic process of creating a short form of an original text. The main goal of an automatic text summarization system is production of a summary which satisfies the user's needs. In this paper, a new model for automatic text summarization is introduced which is based on fuzzy logic system, evolutionary algorithms and cellular learning automata. First, the most important features including word features, similarity measure, and the position and the length of a sentence are extracted. A linear combination of these features shows the importance of each sentence. To calculate similarity measure, a combined method based on artificial bee colony algorithm and cellular learning automata are used. In this method, joint n-grams among sentences are extracted by cellular learning automata and then an artificial bee colony algorithm classifies n-friends in order to extract data and optimize the similarity measure as fitness function. Moreover, a new approach is proposed to adjust the best weights of the text features using particle swarm optimization and genetic algorithm. This method discovers more important and less important text features and then assigns fair weights to them. At last, a fuzzy logic system is used to perform the final scoring. The results of the proposed approach were compared with the other methods including Msword, System19, System21, System28, System31, FSDH, FEOM, NetSum, CRF, SVM, DE, MA-SingleDocSum, Unified Rank and Manifold Ranking using ROUGE-l and ROUGE-2 measures on the DUC2002 dataset. The results show that proposed method outperforms the aforementioned methods.