Search

A novel method for semantic action recognition through learning a pose lexicon is presented in this paper. A pose lexicon comprises a set of semantic poses, a set of visual poses, and a probabilistic mapping between the visual and semantic poses. This paper assumes that both the visual poses and mapping are hidden and proposes a method to simultaneously learn a visual pose model that estimates the likelihood of an observed video frame being generated from hidden visual poses, and a pose lexicon model establishes the probabilistic mapping between the hidden visual poses and the semantic poses parsed from textual instructions. Specifically, the proposed method consists of two-level hidden Markov models. One level represents the alignment between the visual poses and semantic poses. The other level represents a visual pose sequence, and each visual pose is modeled as a Gaussian mixture. An expectation-maximization algorithm is developed to train a pose lexicon. With the learned lexicon, action classification is formulated as a problem of finding the maximum posterior probability of a given sequence of video frames that follows a given sequence of semantic poses, constrained by the most likely visual pose and the alignment sequences. The proposed method was evaluated on MSRC-12, WorkoutSU-10, WorkoutUOW-18, Combined-15, Combined-17, and Combined-50 action datasets using cross-subject, cross-dataset, zero-shot, and seen/unseen protocols. [ABSTRACT FROM AUTHOR]

This article takes a problem-oriented perspective and presents a comprehensive review of transfer-learning methods, both shallow and deep, for cross-dataset visual recognition. Specifically, it categorises the crossdataset recognition into 17 problems based on a set of carefully chosen data and label attributes. Such a problem-oriented taxonomy has allowed us to examine how different transfer-learning approaches tackle each problem and how well each problem has been researched to date. The comprehensive problem-oriented review of the advances in transfer learning with respect to the problem has not only revealed the challenges in transfer learning for visual recognition but also the problems (e.g., 8 of the 17 problems) that have been scarcely studied. This survey not only presents an up-to-date technical review for researchers but also a systematic approach and a reference for a machine-learning practitioner to categorise a real problem and to look up for a possible solution accordingly. [ABSTRACT FROM AUTHOR]

This paper presents a new vision-based method for real-time assessment of upper-body postures of a subject who is sitting in front of a desk studying or operating a computer. Unlike most existing vision-based methods that perform offline assessment from human skeletons extracted from RGB video or depth maps, the proposed method analyses directly single images captured by a webcam in front of the subject without the prone-to-error process of extracting the skeleton data from the images or depth maps. To this end, this paper proposes to assess postures by classifying them into predefined classes, without explicitly measuring the variables required for calculating risk scores. Each class of postures is associated with a configuration of the upper body, and an ergonomics risk score is assigned by following one of the scoring methods, e.g. Rapid Upper Limb Assessment (RULA). A data set of upper-body postures that cover the various scenarios when a subject is sitting in front of a desk as well as some extreme cases when the subject turns away from the desk is collected for evaluating the proposed method quantitatively. The proposed method achieved an on-average accuracy of 99.5% for binary classification (low- vs. high-risk postures), 88.2% for classification of 19 risk levels and 81.5% for classification of 30 risk levels on the data set, and the demo developed based on the method runs in real time on a regular computer. [ABSTRACT FROM AUTHOR]

Surface braiding angle and pitch length are two important parameters for characterizing and evaluating the performance of three-dimensional braided composite preforms. In this paper a new method based on an improved curvature scale space corner detector with adaptive threshold is proposed for measuring these parameters, with applications to three-dimensional, four-directional carbon-fiber braided composite preforms. First, the original image is acquired using a system employing ‘dome light source + CCD camera + circular polarizing filter’. Second, the original image is processed using Lab transform and BM3D (block-matching and 3D filter). Third, the corners are detected using an improved curvature scale space corner detector with adaptive threshold. Finally, the pitch lengths and surface braiding angles are measured from the detected corners. Experimental results show that the proposed method can achieve automatic measurement of the pitch length and surface braiding angle with smaller average errors relative to manual measurements compared with alternative schemes. [ABSTRACT FROM AUTHOR]

Reconstructing magnetic resonance images from undersampled k-space data is a challenging problem. This paper introduces a novel method of image reconstruction from undersampled k-space data based on the concept of singularizing operators and a novel singular k-space model. Exploring the sparsity of an image in the k-space, the singular k-space model (SKM) is proposed in terms of the k-space functions of a singularizing operator. The singularizing operator is constructed by combining basic difference operators. An algorithm is developed to reliably estimate the model parameters from undersampled k-space data. The estimated parameters are then used to recover the missing k-space data through the model, subsequently achieving high-quality reconstruction of the image using inverse Fourier transform. Experiments on physical phantom and real brain MR images have shown that the proposed SKM method constantly outperforms the popular total variation (TV) and the classical zero-filling (ZF) methods regardless of the undersampling rates, the noise levels, and the image structures. For the same objective quality of the reconstructed images, the proposed method requires much less k-space data than the TV method. The SKM method is an effective method for fast MRI reconstruction from the undersampled k-space data. Graphical abstract Two Real Images and their sparsified images by singularizing operator. [ABSTRACT FROM AUTHOR]

This paper proposes three simple, compact yet effective representations of depth sequences, referred to respectively as dynamic depth images (DDI), dynamic depth normal images (DDNI), and dynamic depth motion normal images (DDMNI), for both isolated and continuous action recognition. These dynamic images are constructed from a segmented sequence of depth maps using hierarchical bidirectional rank pooling to effectively capture the spatial-temporal information. Specifically, DDI exploits the dynamics of postures over time, and DDNI and DDMNI exploit the 3-D structural information captured by depth maps. Upon the proposed representations, a convolutional neural network (ConvNet)-based method is developed for action recognition. The image-based representations enable us to fine-tune the existing ConvNet models trained on image data without training a large number of parameters from scratch. The proposed method achieved the state-of-art results on three large datasets, namely, the large-scale continuous gesture recognition dataset (means the Jaccard index 0.4109), the large-scale isolated gesture recognition dataset ( \text59.21\%), and the NTU RGB+D dataset ( \text87.08\% cross-subject and \text84.22\% cross-view) even though only the depth modality was used. [ABSTRACT FROM PUBLISHER]

This paper proposes novel methods for detecting and separating smoke from a single image frame. Specifically, an image formation model is derived based on the atmospheric scattering models. The separation of a frame into quasi-smoke and quasi-background components is formulated as convex optimization that solves a sparse representation problem using dual dictionaries for the smoke and background components, respectively. A novel feature is constructed as a concatenation of the respective sparse coefficients for detection. In addition, a method based on the concept of image matting is developed to separate the true smoke and background components from the smoke detection results. Extensive experiments on detection were conducted and the results showed that the proposed feature significantly outperforms existing features for smoke detection. In particular, the proposed method is able to differentiate smoke from other challenging objects (e.g. fog/haze, cloud, and so on) with similar visual appearance in a gray-scale frame. Experiments on smoke separation also demonstrated that the proposed separation method can effectively estimate/separate the true smoke and background components. [ABSTRACT FROM PUBLISHER]

The emission function associated with the economic-emission dispatch prob- lem contains exponential functions that model the emission pollutants. This paper presents a strategy of solving the economic-emission dispatch problem whereby the ex- ponential function is approximated by a rational function that permits reduction to a standard polynomial optimization problem. This is reformulated as a hierarchy of semidefinite relaxation problems using the moment theory and the resulting SDP prob- lem is solved. Different degrees of rational functional approximation were considered. The approach was tested on the IEEE 30-bus test systems to investigate its effectiveness. Solutions obtained were compared with those from some of the well known evolutionary methods. Results showed that SDP has inherently good convergence property and a lower but comparable diversity property. [ABSTRACT FROM AUTHOR]

Background: Non-proliferative diabetic retinopathy is the early stage of diabetic retinopathy. Automatic detection of non-proliferative diabetic retinopathy is significant for clinical diagnosis, early screening and course progression of patients.Methods: This paper introduces the design and implementation of an automatic system for screening non-proliferative diabetic retinopathy based on color fundus images. Firstly, the fundus structures, including blood vessels, optic disc and macula, are extracted and located, respectively. In particular, a new optic disc localization method using parabolic fitting is proposed based on the physiological structure characteristics of optic disc and blood vessels. Then, early lesions, such as microaneurysms, hemorrhages and hard exudates, are detected based on their respective characteristics. An equivalent optical model simulating human eyes is designed based on the anatomical structure of retina. Main structures and early lesions are reconstructed in the 3D space for better visualization. Finally, the severity of each image is evaluated based on the international criteria of diabetic retinopathy.Results: The system has been tested on public databases and images from hospitals. Experimental results demonstrate that the proposed system achieves high accuracy for main structures and early lesions detection. The results of severity classification for non-proliferative diabetic retinopathy are also accurate and suitable.Conclusions: Our system can assist ophthalmologists for clinical diagnosis, automatic screening and course progression of patients. [ABSTRACT FROM AUTHOR]

Ground parrot vocalisation can be considered as an audio event. Test-based diverse density multiple instance learning (TB-DD-MIL) is proposed for detecting this event in audio files recorded in the field. The proposed method is motivated by the advantages of multiple instance learning from incomplete training data. Spectral features suitable for encoding the vocal source information of the ground parrot vocalization are also investigated. The proposed method was benchmarked against a dataset collected in various environmental conditions and an audio detection evaluation scheme is proposed. The evaluation includes a study on performance of the various vocal source features and comparison with other classification techniques. Experimental results indicated that the most appropriate feature to encode ground parrot calls is the spectral bandwidth and the proposed TB-DD-MIL method outperformed other existing classification methods. [ABSTRACT FROM AUTHOR]

Due to its causal semantics, Bayesian networks (BN) have been widely employed to discover the underlying data relationship in exploratory studies, such as brain research. Despite its success in modeling the probability distribution of variables, BN is naturally a generative model, which is not necessarily discriminative. This may cause the ignorance of subtle but critical network changes that are of investigation values across populations. In this paper, we propose to improve the discriminative power of BN models for continuous variables from two different perspectives. This brings two general discriminative learning frameworks for Gaussian Bayesian networks (GBN). In the first framework, we employ Fisher kernel to bridge the generative models of GBN and the discriminative classifiers of SVMs, and convert the GBN parameter learning to Fisher kernel learning via minimizing a generalization error bound of SVMs. In the second framework, we employ the max-margin criterion and build it directly upon GBN models to explicitly optimize the classification performance of the GBNs. The advantages and disadvantages of the two frameworks are discussed and experimentally compared. Both of them demonstrate strong power in learning discriminative parameters of GBNs for neuroimaging based brain network analysis, as well as maintaining reasonable representation capacity. The contributions of this paper also include a new Directed Acyclic Graph (DAG) constraint with theoretical guarantee to ensure the graph validity of GBN. [ABSTRACT FROM PUBLISHER]

Existing video-based smoke detection methods often rely on the visual features extracted directly from the original frames. In the case of light smoke, the background is still visible and it deteriorates the quality of the features. This paper presents an approach to separating the smoke component from the background such that visual features can be extracted from the smoke component for reliable smoke detection. Specifically, an image is assumed to be a linear blending of a smoke component and a background image. Given a video frame and its background, the estimation of the blending parameter and the actual smoke component can be formulated as an optimization problem. Three methods based on different models for the smoke component are proposed to solve the optimization problem. Experimental results on synthesized and real video data have shown that the proposed approach can effectively separate the smoke component and the smoke detection performance is significantly improved by using the visual features extracted from the smoke component. [ABSTRACT FROM AUTHOR]

This paper examines filtering on a sphere, by first examining the roles of spherical harmonic magnitude and phase. We show that phase is more important than magnitude in determining the structure of a spherical function. We examine the properties of linear phase shifts in the spherical harmonic domain, which suggest a mechanism for constructing finite-impulse-response (FIR) filters. We show that those filters have desirable properties, such as being associative, mapping spherical functions to spherical functions, allowing directional filtering, and being defined by relatively simple equations. We provide examples of the filters for both spherical and manifold data. [ABSTRACT FROM PUBLISHER]

Abstract-It is a common approach for JPEG and MPEG encryption systems to provide higher protection for dc coefficients and less protection for ac coefficients. Some authors have employed a cryptographic encryption algorithm for the dc coefficients and left the ac coefficients to techniques based on random permutation lists which are known to be weak against known-plaintext and chosen-ciphertext attacks. In this paper we show that in block-based DCT, it is possible to recover dc coefficients from ac coefficients with reasonable image quality and show the insecurity of image encryption methods which rely on the encryption of dc values using a cryptoalgorithm. The method proposed in this paper combines dc recovery from ac coefficients and the fact that ac coefficients can be recovered using a chosen ciphertext attack. We demonstrate that a method proposed by Tang to encrypt and decrypt MPEG video can be completely broken. [ABSTRACT FROM AUTHOR]

Presents two algorithms for removing the cross-talk effect in complementary metal-oxide-semiconductor (CMOS) sensor based color-imaging systems. Parts of a typical digital color imaging system with one-sensor CMOS imager; Factors that contribute to the cross-talk; Formulation of the cross-talk compensation problem.

Proposes a multiresolution form of the singular value decomposition (SVD). Notation and conventions; How the multiresolution SVD may be constructed; Complexity analysis; Interpretation of decomposition; Calculations for the Markov-1 model.