Search

Linear discriminant analysis (LDA) has been widely used as the technique of feature exaction. However, LDA may be invalid to address the data from different domains. The reasons are as follows: 1) the distribution discrepancy of data may disturb the linear transformation matrix so that it cannot extract the most discriminative feature and 2) the original design of LDA does not consider the unlabeled data so that the unlabeled data cannot take part in the training process for further improving the performance of LDA. To address these problems, in this brief, we propose a novel transferable LDA (TLDA) method to extend LDA into the scenario in which the data have different probability distributions. The whole learning process of TLDA is driven by the philosophy that the data from the same subspace have a low-rank structure. The matrix rank in TLDA is the key learning criterion to conduct local and global linear transformations for restoring the low-rank structure of data from different distributions and enlarging the distances among different subspaces. In doing so, the variations of distribution discrepancy within the same subspace can be reduced, i.e., data can be aligned well and the maximally separated structure can be achieved for the data from different subspaces. A simple projected subgradient-based method is proposed to optimize the objective of TLDA, and a strict theory proof is provided to guarantee a quick convergence. The experimental evaluation on public data sets demonstrates that our TLDA can achieve better classification performance and outperform the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Video captioning, in essential, is a complex natural process, which is affected by various uncertainties stemming from video content, subjective judgment, and so on. In this paper, we build on the recent progress in using encoder-decoder framework for video captioning and address what we find to be a critical deficiency of the existing methods that most of the decoders propagate deterministic hidden states. Such complex uncertainty cannot be modeled efficiently by the deterministic models. In this paper, we propose a generative approach, referred to as multimodal stochastic recurrent neural networks (MS-RNNs), which models the uncertainty observed in the data using latent stochastic variables. Therefore, MS-RNN can improve the performance of video captioning and generate multiple sentences to describe a video considering different random factors. Specifically, a multimodal long short-term memory (LSTM) is first proposed to interact with both visual and textual features to capture a high-level representation. Then, a backward stochastic LSTM is proposed to support uncertainty propagation by introducing latent variables. Experimental results on the challenging data sets, microsoft video description and microsoft research video-to-text, show that our proposed MS-RNN approach outperforms the state-of-the-art video captioning benchmarks. [ABSTRACT FROM AUTHOR]

Most existing object detection algorithms are trained based upon a set of fully annotated object regions or bounding boxes, which are typically labor-intensive. On the contrary, nowadays there is a significant amount of image-level annotations cheaply available on the Internet. It is hence a natural thought to explore such “weak” supervision to benefit the training of object detectors. In this paper, we propose a novel scheme to perform weakly supervised object localization, termed object-specific pixel gradient (OPG). The OPG is trained by using image-level annotations alone, which performs in an iterative manner to localize potential objects in a given image robustly and efficiently. In particular, we first extract an OPG map to reveal the contributions of individual pixels to a given object category, upon which an iterative mining scheme is further introduced to extract instances or components of this object. Moreover, a novel average and max pooling layer is introduced to improve the localization accuracy. In the task of weakly supervised object localization, the OPG achieves a state-of-the-art 44.5% top-5 error on ILSVRC 2013, which outperforms competing methods, including Oquabet al.and region-based convolutional neural networks on the Pascal VOC 2012, with gains of 2.6% and 2.3%, respectively. In the task of object detection, OPG achieves a comparable performance of 27.0% mean average precision on Pascal VOC 2007. In all experiments, the OPG only adopts the off-the-shelf pretrained CNN model, without using any object proposals. Therefore, it also significantly improves the detection speed, i.e., achieving three times faster compared with the state-of-the-art method. [ABSTRACT FROM AUTHOR]

Conventional convolutional neural networks use either a linear or a nonlinear filter to extract features from an image patch (region) of spatial size $ H\times W $ (typically, $ H $ is small and is equal to $ W$ , e.g., $ H $ is 5 or 7). Generally, the size of the filter is equal to the size $ H\times W $ of the input patch. We argue that the representational ability of equal-size strategy is not strong enough. To overcome the drawback, we propose to use subpatch filter whose spatial size $ h\times w $ is smaller than $ H\times W $. The proposed subpatch filter consists of two subsequent filters. The first one is a linear filter of spatial size $ h\times w $ and is aimed at extracting features from spatial domain. The second one is of spatial size $ 1\times 1 $ and is used for strengthening the connection between different input feature channels and for reducing the number of parameters. The subpatch filter convolves with the input patch and the resulting network is called a subpatch network. Taking the output of one subpatch network as input, we further repeat constructing subpatch networks until the output contains only one neuron in spatial domain. These subpatch networks form a new network called the cascaded subpatch network (CSNet). The feature layer generated by CSNet is called the csconv layer. For the whole input image, we construct a deep neural network by stacking a sequence of csconv layers. Experimental results on five benchmark data sets demonstrate the effectiveness and compactness of the proposed CSNet. For example, our CSNet reaches a test error of 5.68% on the CIFAR10 data set without model averaging. To the best of our knowledge, this is the best result ever obtained on the CIFAR10 data set. [ABSTRACT FROM AUTHOR]

It is observed that distinct words in a given document have either strong or weak ability in delivering facts (i.e., the objective sense) or expressing opinions (i.e., the subjective sense) depending on the topics they associate with. Motivated by the intuitive assumption that different words have varying degree of discriminative power in delivering the objective sense or the subjective sense with respect to their assigned topics, a model named as i dentified o bjective– s ubjective latent Dirichlet allocation (LDA) ( i osLDA) is proposed in this paper. In the i osLDA model, the simple Pólya urn model adopted in traditional topic models is modified by incorporating it with a probabilistic generative process, in which the novel “Bag-of-Discriminative-Words” (BoDW) representation for the documents is obtained; each document has two different BoDW representations with regard to objective and subjective senses, respectively, which are employed in the joint objective and subjective classification instead of the traditional Bag-of-Topics representation. The experiments reported on documents and images demonstrate that: 1) the BoDW representation is more predictive than the traditional ones; 2) i osLDA boosts the performance of topic modeling via the joint discovery of latent topics and the different objective and subjective power hidden in every word; and 3) i osLDA has lower computational complexity than supervised LDA, especially under an increasing number of topics. [ABSTRACT FROM PUBLISHER]

Eye movements in the case of freely viewing natural scenes are believed to be guided by local contrast, global contrast, and top-down visual factors. Although a lot of previous works have explored these three saliency cues for several years, there still exists much room for improvement on how to model them and integrate them effectively. This paper proposes a novel computation model to predict eye fixations, which adopts a multiresolution convolutional neural network (Mr-CNN) to infer these three types of saliency cues from raw image data simultaneously. The proposed Mr-CNN is trained directly from fixation and nonfixation pixels with multiresolution input image regions with different contexts. It utilizes image pixels as inputs and eye fixation points as labels. Then, both the local and global contrasts are learned by fusing information in multiple contexts. Meanwhile, various top-down factors are learned in higher layers. Finally, optimal combination of top-down factors and bottom-up contrasts can be learned to predict eye fixations. The proposed approach significantly outperforms the state-of-the-art methods on several publically available benchmark databases, demonstrating the superiority of Mr-CNN. We also apply our method to the RGB-D image saliency detection problem. Through learning saliency cues induced by depth and RGB information on pixel level jointly and their interactions, our model achieves better performance on predicting eye fixations in RGB-D images. [ABSTRACT FROM AUTHOR]

Matching people across nonoverlapping cameras, also known as person re-identification, is an important and challenging research topic. Despite its great demand in many crucial applications such as surveillance, person re-identification is still far from being solved. Due to drastic view changes, even the same person may look quite dissimilar in different cameras. Illumination and pose variations further aggravate this discrepancy. To this end, various feature descriptors have been designed for improving the matching accuracy. Since different features encode information from different aspects, in this paper, we propose to effectively leverage multiple off-the-shelf features via multi-hypergraph fusion. A hypergraph captures not only pairwise but also high-order relationships among the subjects being matched. In addition, different from conventional approaches in which the matching is achieved by computing the pairwise distance or similarity between a probe and a gallery subject, the similarities between the probe and all gallery subjects are learned jointly via hypergraph optimization. Experiments on popular data sets demonstrate the effectiveness of the proposed method, and a superior performance is achieved as compared with the most recent state-of-the-arts. [ABSTRACT FROM PUBLISHER]

In this paper, we propose a biologically inspired appearance model for robust visual tracking. Motivated in part by the success of the hierarchical organization of the primary visual cortex (area V1), we establish an architecture consisting of five layers: whitening, rectification, normalization, coding, and pooling. The first three layers stem from the models developed for object recognition. In this paper, our attention focuses on the coding and pooling layers. In particular, we use a discriminative sparse coding method in the coding layer along with spatial pyramid representation in the pooling layer, which makes it easier to distinguish the target to be tracked from its background in the presence of appearance variations. An extensive experimental study shows that the proposed method has higher tracking accuracy than several state-of-the-art trackers. [ABSTRACT FROM AUTHOR]

Visual feature learning, which aims to construct an effective feature representation for visual data, has a wide range of applications in computer vision. It is often posed as a problem of nonnegative matrix factorization (NMF), which constructs a linear representation for the data. Although NMF is typically parallelized for efficiency, traditional parallelization methods suffer from either an expensive computation or a high runtime memory usage. To alleviate this problem, we propose a parallel NMF method called alternating least square block decomposition (ALSD), which efficiently solves a set of conditionally independent optimization subproblems based on a highly parallelized fine-grained grid-based blockwise matrix decomposition. By assigning each block optimization subproblem to an individual computing node, ALSD can be effectively implemented in a MapReduce-based Hadoop framework. In order to cope with dynamically varying visual data, we further present an incremental version of ALSD, which is able to incrementally update the NMF solution with a low computational cost. Experimental results demonstrate the efficiency and scalability of the proposed methods as well as their applications to image clustering and image retrieval. [ABSTRACT FROM PUBLISHER]

Salient object detection increasingly receives attention as an important component or step in several pattern recognition and image processing tasks. Although a variety of powerful saliency models have been intensively proposed, they usually involve heavy feature (or model) engineering based on priors (or assumptions) about the properties of objects and backgrounds. Inspired by the effectiveness of recently developed feature learning, we provide a novel deep image saliency computing (DISC) framework for fine-grained image saliency computing. In particular, we model the image saliency from both the coarse-and fine-level observations, and utilize the deep convolutional neural network (CNN) to learn the saliency representation in a progressive manner. In particular, our saliency model is built upon two stacked CNNs. The first CNN generates a coarse-level saliency map by taking the overall image as the input, roughly identifying saliency regions in the global context. Furthermore, we integrate superpixel-based local context information in the first CNN to refine the coarse-level saliency map. Guided by the coarse saliency map, the second CNN focuses on the local context to produce fine-grained and accurate saliency map while preserving object details. For a testing image, the two CNNs collaboratively conduct the saliency computing in one shot. Our DISC framework is capable of uniformly highlighting the objects of interest from complex background while preserving well object details. Extensive experiments on several standard benchmarks suggest that DISC outperforms other state-of-the-art methods and it also generalizes well across data sets without additional training. The executable version of DISC is available online: http://vision.sysu.edu.cn/projects/DISC. [ABSTRACT FROM PUBLISHER]

Visual saliency is one of the most noteworthy perceptual abilities of human vision. Recent progress in cognitive psychology suggests that: 1) visual saliency analysis is mainly completed by the bottom-up mechanism consisting of feedforward low-level processing in primary visual cortex (area V1) and 2) color interacts with spatial cues and is influenced by the neighborhood context, and thus it plays an important role in a visual saliency analysis. From a computational perspective, the most existing saliency modeling approaches exploit multiple independent visual cues, irrespective of their interactions (or are not computed explicitly), and ignore contextual influences induced by neighboring colors. In addition, the use of color is often underestimated in the visual saliency analysis. In this paper, we propose a simple yet effective color saliency model that considers color as the only visual cue and mimics the color processing in V1. Our approach uses region-/boundary-defined color features with spatiochromatic filtering by considering local color-orientation interactions, therefore captures homogeneous color elements, subtle textures within the object and the overall salient object from the color image. To account for color contextual influences, we present a divisive normalization method for chromatic stimuli through the pooling of contrary/complementary color units. We further define a color perceptual metric over the entire scene to produce saliency maps for color regions and color boundaries individually. These maps are finally globally integrated into a one single saliency map. The final saliency map is produced by Gaussian blurring for robustness. We evaluate the proposed method on both synthetic stimuli and several benchmark saliency data sets from the visual saliency analysis to salient object detection. The experimental results demonstrate that the use of color as a unique visual cue achieves competitive results on par with or better than 12 state-of-the-art approaches. [ABSTRACT FROM PUBLISHER]

Spatial pyramid matching is a standard architecture for categorical image retrieval. However, its performance is largely limited by the prespecified rectangular spatial regions when pooling local descriptors. In this paper, we propose to learn object-shaped and directional receptive fields for image categorization. In particular, different objects in an image are seamlessly constructed by superpixels, while the direction captures human gaze shifting path. By generating a number of superpixels in each image, we construct graphlets to describe different objects. They function as the object-shaped receptive fields for image comparison. Due to the huge number of graphlets in an image, a saliency-guided graphlet selection algorithm is proposed. A manifold embedding algorithm encodes graphlets with the semantics of training image tags. Then, we derive a manifold propagation to calculate the postembedding graphlets by leveraging visual saliency maps. The sequentially propagated graphlets constitute a path that mimics human gaze shifting. Finally, we use the learned graphlet path as receptive fields for local image descriptor pooling. The local descriptors from similar receptive fields of pairwise images more significantly contribute to the final image kernel. Thorough experiments demonstrate the advantage of our approach. [ABSTRACT FROM PUBLISHER]

We investigate the tracking of 2-D human poses in a video stream to determine the spatial configuration of body parts in each frame, but this is not a trivial task because people may wear different kinds of clothing and may move very quickly and unpredictably. The technology of pose estimation is typically applied, but it ignores the temporal context and cannot provide smooth, reliable tracking results. Therefore, we develop a tracking and estimation integrated model (TEIM) to fully exploit temporal information by integrating pose estimation with visual tracking. However, joint parsing of multiple articulated parts over time is difficult, because a full model with edges capturing all pairwise relationships within and between frames is loopy and intractable. In previous models, approximate inference was usually resorted to, but it cannot promise good results and the computational cost is large. We overcome these problems by exploring the idea of divide and conquer, which decomposes the full model into two much simpler tractable submodels. In addition, a novel two-step iteration strategy is proposed to efficiently conquer the joint parsing problem. Algorithmically, we design TEIM very carefully so that: 1) it enables pose estimation and visual tracking to compensate for each other to achieve desirable tracking results; 2) it is able to deal with the problem of tracking loss; and 3) it only needs past information and is capable of tracking online. Experiments are conducted on two public data sets in the wild with ground truth layout annotations, and the experimental results indicate the effectiveness of the proposed TEIM framework. [ABSTRACT FROM PUBLISHER]

This paper investigates how to blindly evaluate the visual quality of an image by learning rules from linguistic descriptions. Extensive psychological evidence shows that humans prefer to conduct evaluations qualitatively rather than numerically. The qualitative evaluations are then converted into the numerical scores to fairly benchmark objective image quality assessment (IQA) metrics. Recently, lots of learning-based IQA models are proposed by analyzing the mapping from the images to numerical ratings. However, the learnt mapping can hardly be accurate enough because some information has been lost in such an irreversible conversion from the linguistic descriptions to numerical scores. In this paper, we propose a blind IQA model, which learns qualitative evaluations directly and outputs numerical scores for general utilization and fair comparison. Images are represented by natural scene statistics features. A discriminative deep model is trained to classify the features into five grades, corresponding to five explicit mental concepts, i.e., excellent, good, fair, poor, and bad. A newly designed quality pooling is then applied to convert the qualitative labels into scores. The classification framework is not only much more natural than the regression-based models, but also robust to the small sample size problem. Thorough experiments are conducted on popular databases to verify the model’s effectiveness, efficiency, and robustness. [ABSTRACT FROM AUTHOR]

Regular machine learning and data mining techniques study the training data for future inferences under a major assumption that the future data are within the same feature space or have the same distribution as the training data. However, due to the limited availability of human labeled training data, training data that stay in the same feature space or have the same distribution as the future data cannot be guaranteed to be sufficient enough to avoid the over-fitting problem. In real-world applications, apart from data in the target domain, related data in a different domain can also be included to expand the availability of our prior knowledge about the target future data. Transfer learning addresses such cross-domain learning problems by extracting useful information from data in a related domain and transferring them for being used in target tasks. In recent years, with transfer learning being applied to visual categorization, some typical problems, e.g., view divergence in action recognition tasks and concept drifting in image classification tasks, can be efficiently solved. In this paper, we survey state-of-the-art transfer learning algorithms in visual categorization applications, such as object recognition, image classification, and human action recognition. [ABSTRACT FROM PUBLISHER]