Search

This research explores indicators and methods for an enterprise to measure and evaluate user satisfaction with enterprise social media for knowledge management. This paper presents qualitative indicators, including three service levels of enterprise social media for knowledge management (KM) from a techno-social perspective. This research puts forward a synthetic evaluation model mixed with linguistic variables, consistent fuzzy preference relations (CFPR) and cloud model for measuring and evaluating user satisfaction. The synthetic evaluation model can transform linguistic variables into quantitative data to obtain user satisfaction levels and determine the distance between the expected satisfaction level and actual performance. This research can help an enterprise to improve the service ability of its social media to meet users’ requirements for knowledge management.

Depth sensing with single-shot structured light is widely employed for depth measurement of dynamic scenes. For conventional single-shot structured light techniques, the projected pattern is coded with regards to color, shape, and intensity, aiming at assigning each pixel a unique label. However, the features of the projected pattern are extremely susceptible to the surface texture of the measured object, resulting in errors during performing decoding. In this study, a robust and simple binary pattern is proposed for high-accuracy depth sensing. Only two gray values are utilized in the designed pattern, which is insensitive to environmental noise. The pseudo two-dimensional De Bruijn sequence coding method is employed to reduce the size of the decoding window for robust matching. Short slashes are adopted as features of the pattern that greatly simplifies the feature extraction and pattern decoding. Compared with the Kinect and time-of-flight (TOF) depth cameras, our proposed method can acquire high-precision depth in quantitative and qualitative experiments.

Currently, video text spotting tasks usually fall into the four-staged pipeline: detecting text regions in individual images, recognizing localized text regions frame-wisely, tracking text streams and post-processing to generate final results. However, they may suffer from the huge computational cost as well as sub-optimal results due to the interferences of low-quality text and the none-trainable pipeline strategy. In this article, we propose a fast and robust end-to-end video text spotting framework named FREE by only recognizing the localized text stream one-time instead of frame-wise recognition. Specifically, FREE first employs a well-designed spatial-temporal detector that learns text locations among video frames. Then a novel text recommender is developed to select the highest-quality text from text streams for recognizing. Here, the recommender is implemented by assembling text tracking, quality scoring and recognition into a trainable module. It not only avoids the interferences from the low-quality text but also dramatically speeds up the video text spotting. FREE unites the detector and recommender into a whole framework, and helps achieve global optimization. Besides, we collect a large scale video text dataset for promoting the video text spotting community, containing 100 videos from 21 real-life scenarios. Extensive experiments on public benchmarks show our method greatly speeds up the text spotting process, and also achieves the remarkable state-of-the-art.

Structured light techniques are widely used in the depth measurement. In this paper, we propose a colored speckle method for high-accuracy depth sensing. The projected pattern is a composite of the discrete phase information and the colored De Bruijn sequence information, which can prove the feasibility, in practice, and the robustness to the environment. A Gabor filter is used to estimate the wrapped phase. To alleviate the phase ambiguity of conventional measurement methods, the period of the phase is distributed in two channels of the colored pattern by the De Bruijn sequence decoding. A morphological closure is adopted to determine the De Bruijn code by the phase neighborhoods back and forth. The proposed method is suitable for a dynamic scene. The quantitative and qualitative experiments are conducted, in practice, to evaluate the performance of the proposed method. The results of the experiments prove that not only can our proposed method achieve high accuracy and dense depth map, but it is also suitable for color object and dynamic scenes indoors.

This paper proposes a segregated temporal assembly recurrent (STAR) network for weakly-supervised multiple action detection. The model learns from untrimmed videos with only supervision of video-level labels and makes prediction of intervals of multiple actions. Specifically, we first assemble video clips according to class labels by an attention mechanism that learns class-variable attention weights and thus helps the noise relieving from background or other actions. Secondly, we build temporal relationship between actions by feeding the assembled features into an enhanced recurrent neural network. Finally, we transform the output of recurrent neural network into the corresponding action distribution. In order to generate more precise temporal proposals, we design a score term called segregated temporal gradient-weighted class activation mapping (ST-GradCAM) fused with attention weights. Experiments on THUMOS'14 and ActivityNet1.3 datasets show that our approach outperforms the state-of-the-art weakly-supervised method, and performs at par with the fully-supervised counterparts., Comment: Accepted to Proc. AAAI Conference on Artificial Intelligence 2019

The methods of structured light and deep learning are widely used in artificial vision to acquire a depth map of real-world scenes. In this paper, we propose a novel method of combining structured light and deep learning stereo matching to calculate the depth. To combat the problems with textureless areas of stereo matching, a pair of left and right side images with abundant structured light information is adopted to acquire a coarse depth map by unsupervised CNN networks. The coarse depth map is used for phase unwrapping, which is a bottleneck in the phase-based structured light method. Then, a fine and accurate depth map is obtained by phase matching. Compared with the traditional structured light method, the proposed method performs better on occluded and textureless areas. To evaluate the performance of our proposed method, an experimental platform is established and several experiments are conducted. Quantitative and qualitative experiments demonstrate that the proposed method can generate a high precision depth and relieve the occlusion in the structured light system.

We address a challenging problem: recognizing multiple text sequences from an image by pure end-to-end learning. It is twofold: 1) Multiple text sequences recognition. Each image may contain multiple text sequences of different content, location and orientation, we try to recognize all these texts in the image. 2) Pure end-to-end (PEE) learning. We solve the problem in a pure end-to-end learning way where each training image is labeled by only text transcripts of the contained sequences, without any geometric annotations. Most existing works recognize multiple text sequences from an image in a non-end-to-end (NEE) or quasi-end-to-end (QEE) way, in which each image is trained with both text transcripts and text locations. Only recently, a PEE method was proposed to recognize text sequences from an image where the text sequence was split to several lines in the image. However, it cannot be directly applied to recognizing multiple text sequences from an image. So in this paper, we propose a pure end-to-end learning method to recognize multiple text sequences from an image. Our method directly learns the probability distribution of multiple sequences conditioned on each input image, and outputs multiple text transcripts with a well-designed decoding strategy. To evaluate the proposed method, we construct several datasets mainly based on an existing public dataset and two real application scenarios. Experimental results show that the proposed method can effectively recognize multiple text sequences from images, and outperforms CTC-based and attention-based baseline methods.

Text recognition is a popular topic for its broad applications. In this work, we excavate the implicit task, character counting within the traditional text recognition, without additional labor annotation cost. The implicit task plays as an auxiliary branch for complementing the sequential recognition. We design a two-branch reciprocal feature learning framework in order to adequately utilize the features from both the tasks. Through exploiting the complementary effect between explicit and implicit tasks, the feature is reliably enhanced. Extensive experiments on 7 benchmarks show the advantages of the proposed methods in both text recognition and the new-built character counting tasks. In addition, it is convenient yet effective to equip with variable networks and tasks. We offer abundant ablation studies, generalizing experiments with deeper understanding on the tasks. Code is available.

The basic principle of the patch-matching based style transfer is to substitute the patches of the content image feature maps by the closest patches from the style image feature maps. Since the finite features harvested from one single aesthetic style image are inadequate to represent the rich textures of the content natural image, existing techniques treat the full-channel style feature patches as simple signal tensors and create new style feature patches via signal-level fusion. In this paper, we propose a channel-grouping based patch swap technique to group the style feature maps into surface and texture channels, and the new features are created by the combination of these two groups, which can be regarded as a semantic-level fusion of the raw style features. Experimental results demonstrate that the proposed method outperforms the existing techniques in providing more style-consistent textures while keeping the content fidelity.

Epilepsy is a brain disorder caused by abnormal discharges of neurons in brain. It is one of the most commonly studied disorders in neurology. The research of epilepsy electroencephalogram (EEG) has become a hot research topic. We find that in epilepsy EEG detection task, many previous methods focused on directly collecting the data of each channel, but these methods seldom analyse relationships between signals. Therefore, we propose the Epilepsy EEG Graph Convolutional Network EGCN, which makes full use of correlations between channels to deeply mine data information. We specifically design 5-layer graph convolutional network structure for classification of healthy and epileptic patients. The method is applied to public data set (Boon and CHB-MIT) to establish a reasonable classification model. And we compare it with some advanced algorithms. The experimental results show that the E-GCN method is superior to many existing methods in classification accuracy. In brief, the E-GCN method can be effectively used in classification and detection for epilepsy. This provides new ideas for colleagues, who study epilepsy EEG. In addition, this also provides richer experience for diagnosis of epilepsy.

Scene text recognition (STR) is still a hot research topic in computer vision field due to its various applications. Existing works mainly focus on learning a general model with a huge number of synthetic text images to recognize unconstrained scene texts, and have achieved substantial progress. However, these methods are not quite applicable in many real-world scenarios where 1) high recognition accuracy is required, while 2) labeled samples are lacked. To tackle this challenging problem, this paper proposes a few-shot adversarial sequence domain adaptation (FASDA) approach to build sequence adaptation between the synthetic source domain (with many synthetic labeled samples) and a specific target domain (with only some or a few real labeled samples). This is done by simultaneously learning each character's feature representation with an attention mechanism and establishing the corresponding character-level latent subspace with adversarial learning. Our approach can maximize the character-level confusion between the source domain and the target domain, thus achieves the sequence-level adaptation with even a small number of labeled samples in the target domain. Extensive experiments on various datasets show that our method significantly outperforms the finetuning scheme, and obtains comparable performance to the state-of-the-art STR methods., 8 pages, 6 figures

Since real-world ubiquitous documents (e.g., invoices, tickets, resumes and leaflets) contain rich information, automatic document image understanding has become a hot topic. Most existing works decouple the problem into two separate tasks, (1) text reading for detecting and recognizing texts in images and (2) information extraction for analyzing and extracting key elements from previously extracted plain text. However, they mainly focus on improving information extraction task, while neglecting the fact that text reading and information extraction are mutually correlated. In this paper, we propose a unified end-to-end text reading and information extraction network, where the two tasks can reinforce each other. Specifically, the multimodal visual and textual features of text reading are fused for information extraction and in turn, the semantics in information extraction contribute to the optimization of text reading. On three real-world datasets with diverse document images (from fixed layout to variable layout, from structured text to semi-structured text), our proposed method significantly outperforms the state-of-the-art methods in both efficiency and accuracy., Accepted to ACM MM2020. Code is available at https://davar-lab.github.io/publication.html or https://github.com/hikopensource/DAVAR-Lab-OCR

Multi-spectral imaging (MSI) is a novel non-invasive tool for visualizing the entire span of the eye, from the internal limiting membrane to the choroid. However, spatial misalignments can be frequently observed in sequential MSI images because the eye saccade movement is usually faster than the MSI image acquisition speed. Therefore, registering MSI images is necessary for computer-based analysis of retinal degeneration via MSI. In this paper, we propose an early deep learning framework for achieving an accurate registration of MSI images in a group-wise fashion. The framework contains three parts: a template construction based on principal component analysis, a deformation field calculation, and a spatial transformation. The framework is uniquely capable of resolving two key challenges, i.e., the “multi-modal” characteristics in MSI images for the acquisition with different spectra and the requirement of joint registration of the sequential images. Our experimental results demonstrate the superior performance of our framework compared to several representative state-of-the-art techniques in both speed and accuracy.

Unsafe driving behaviors are the leading causes of truck crashes. Therefore, an enhanced understanding of truck drivers' unsafe driving behaviors is of considerable significance for preventing truck crashes. However, previous studies have rarely encompassed proactive factors such as safety management. Therefore, a classification framework for truck drivers' unsafe driving behaviors was established according to a survey of 2000 truck drivers using four machine learning models (CART, RT, AdaBoost, and GBDT). The classification framework included six first-level input dimensions, 51 s-level input indicators, covering both objective and proactive factors. Nine types of unsafe driving behaviors were determined as outputs. Unique risk factors associated with each of nine unsafe driving behaviors were identified. The results showed that the model's predictive performance varies with different driving behaviors (Classification Accuracy ranges from 0.64 to 0.95, F1-score ranges from 0.52 to 0.72), which was caused by different formation mechanisms of different driving behaviors. Similarly, the results related to factor importance for different driving behaviors were also significantly different, regardless of the first-level and second-level factors. Furthermore, the correlation analysis and OR value strengthened the interpretability of the factor importance, revealing possible reasons for the differences between various driving behaviors.

Traditional sparse representation-(SR) based inverse synthetic aperture radar (ISAR) imaging schemes can achieve significant performance, but they suffer from high costs of memory and computational complexity, because the SR of a 2-D image is converted into that of a 1-D vector. Instead of memory consuming vector operations, we propose a fast ISAR imaging algorithm, where the decomposition and reconstruction of a 2-D scene is implemented using matrix operations directly. Besides the spatial sparsity of a scene, its structural sparsity is presented using the range profile of the scene, where both can be used to enhance sparsity exploitation during image reconstruction. Also, benefitting from the structural sparsity of the range profile, the target energy can be accumulated during the process, which further improves performance. Compared to available SR-based ISAR imaging algorithms, the proposed algorithm reduces both memory costs and computational complexity significantly, which is proven using simulated and real data.

In structured light illumination (SLI) systems, multi-shot fringe patterns can reach higher precision than a single-shot fringe pattern. However, multi-shot methods are not suitable for dynamic scenes while single-shot ones are limited in the measurement accuracy. In this paper, a novel single-shot depth sensing method with frequency-division multiplexing (FDM) framework is proposed. To achieve a simultaneous casting, two fringe patterns with coprime periods are modulated into a single pattern. The method of fringe pattern extraction is similar to the demodulation in communication systems. The Gabor filter is adopted to get the phase information in the pattern, and the coprime theorem is used to solve the phase ambiguity. Quantitative and qualitative evaluations have proved that our method achieves higher accuracy in depth sensing compared with the Kinect v1 and ToF camera. In addition, benefiting from the single-shot pattern, our method is suitable for dynamic scenes.

Many approaches have recently been proposed to detect irregular scene text and achieved promising results. However, their localization results may not well satisfy the following text recognition part mainly because of two reasons: 1) recognizing arbitrary shaped text is still a challenging task, and 2) prevalent non-trainable pipeline strategies between text detection and text recognition will lead to suboptimal performances. To handle this incompatibility problem, in this paper we propose an end-to-end trainable text spotting approach named Text Perceptron. Concretely, Text Perceptron first employs an efficient segmentation-based text detector that learns the latent text reading order and boundary information. Then a novel Shape Transform Module (abbr. STM) is designed to transform the detected feature regions into regular morphologies without extra parameters. It unites text detection and the following recognition part into a whole framework, and helps the whole network achieve global optimization. Experiments show that our method achieves competitive performance on two standard text benchmarks, i.e., ICDAR 2013 and ICDAR 2015, and also obviously outperforms existing methods on irregular text benchmarks SCUT-CTW1500 and Total-Text., Accepted by AAAI2020. Code is available at https://davar-lab.github.io/publication.html or https://github.com/hikopensource/DAVAR-Lab-OCR

Hyperspectral images (HSIs) have far more spectral bands than conventional RGB images. The abundant spectral information provides very useful clues for the followup applications, such as classification and anomaly detection. How to extract discriminant features from HSIs is very important. In this work, we propose a novel spatial-spectral features extraction method for HSI classification by Multi-Scale Depthwise Separable Convolutional Neural Network (MDSCNN). This new model consists of a multi-scale atrous convolution module and two bottleneck residual units, which greatly increase the width and depth of the network. In addition, we use depthwise separable convolution instead of traditional 2D or 3D convolution to extract spatial and spectral features. Furthermore, considering classification accuracy can benifit from multi-scale information, we introduce atrous convolution with different dilation rates parallelly to extract more discriminant features of HSIs for classification. Experiments on three standard datasets show that the proposed MDSCNN has got the state-of-the-art accuracy among all compared methods.

Discriminative facial parts are essential for facial expression recognition (FER) tasks because of small inter-class differences and large intra-class variations in expression images. Existing methods localize discriminative regions with the aid of extra facial landmarks, such as action units (AU). However, it consumes a lot of manpower in manually labeling. To address this problem, in this paper, we propose an advanced facial attention based convolutional neural network (FA-CNN) for 2D+3D FER. The main contribution of FA-CNN is the facial attention mechanism, which enables the network to localize the discriminative regions automatically from multi-modality expression images without dense landmark annotations. Experimental results conducted on BU-3DFE demonstrate that FA-CNN achieves state-of-the-art performance comparing with the existing 2D+3D FER techniques, and the discriminative facial parts estimated by the facial attention mechanism are highly interpretable and consistent with human perception.

Temporal action localization is an important yet challenging research topic due to its various applications. Since the frame-level or segment-level annotations of untrimmed videos require amounts of labor expenditure, studies on the weakly-supervised action detection have been springing up. However, most of existing frameworks rely on Class Activation Sequence (CAS) to localize actions by minimizing the video-level classification loss, which exploits the most discriminative parts of actions but ignores the minor regions. In this paper, we propose a novel weakly-supervised framework by adversarial learning of two modules for eliminating such demerits. Specifically, the first module is designed as a well-designed Seeded Sequence Growing (SSG) Network for progressively extending seed regions (namely the highly reliable regions initialized by a CAS-based framework) to their expected boundaries. The second module is a specific classifier for mining trivial or incomplete action regions, which is trained on the shared features after erasing the seeded regions activated by SSG. In this way, a whole network composed of these two modules can be trained in an adversarial manner. The goal of the adversary is to mine features that are difficult for the action classifier. That is, erasion from SSG will force the classifier to discover minor or even new action regions on the input feature sequence, and the classifier will drive the seeds to grow, alternately. At last, we could obtain the action locations and categories from the well-trained SSG and the classifier. Extensive experiments on two public benchmarks THUMOS'14 and ActivityNet1.3 demonstrate the impressive performance of our proposed method compared with the state-of-the-arts., To be appeared in ACM MM2019

Objective Electroencephalogram (EEG) based brain-computer interfaces (BCI) in motor imagery (MI) have developed rapidly in recent years. A reliable feature extraction method is essential because of a low signal-to-noise ratio (SNR) and time-dependent covariates of EEG signals. Because of efficient application in various fields, deep learning has been adopted in EEG signal processing and has obtained competitive results compared with the traditional methods. However, designing and training an end-to-end network to fully extract potential features from EEG signals remains a challenge in MI. Approach In this study, we propose a parallel multiscale filter bank convolutional neural network (MSFBCNN) for MI classification. We introduce a layered end-to-end network structure, in which a feature-extraction network is used to extract temporal and spatial features. To enhance the transfer learning ability, we propose a network initialization and fine-tuning strategy to train an individual model for inter-subject classification on small datasets. We compare our MSFBCNN with the state-of-the-art approaches on open datasets. Results The proposed method has a higher accuracy than the baselines in intra-subject classification. In addition, the transfer learning experiments indicate that our network can build an individual model and obtain acceptable results in inter-subject classification. The results suggest that the proposed network has superior performance, robustness, and transfer learning ability.

Fine-grained visual categorization (FGVC) has attracted extensive attention in recent years. The general pipeline of current FGVC techniques is to 1) locate the discriminative regions; 2) extract features from each region independently; and 3) feed the integrated features to a classifier. In this paper, we re-investigate the pipeline from the view of human visual recognition mechanisms. The perceiving of discriminative regions is a temporal processing by the human visual system (HVS) via the attention-shift mechanism. However, the existing independent feature extracting and one-pass feeding strategy ignore the inherent semantic relationships among discriminative regions, and thus is improper to model the attention-shift process properly. Therefore, in this paper, we propose a novel end-to-end FGVC network structure named Attention-Shift based Deep Neural Network (AS-DNN) to locate the discriminative regions automatically and encode the semantic correlations iteratively. AS-DNN consists of two channels: 1) the global perception channel C glb and 2) the attention-shift channel C sft , simulating the global perception and the attention-shift mechanism, respectively. Experimental results show that AS-DNN achieves state-of-the-art performances by outperforming both the CNN-based weakly or strongly-supervised FGVC algorithms on several widely-used fine-grained datasets, and the visualization of attention regions exhibit that the proposed method can locate the discriminative regions robustly in complex backgrounds and postures.

This study reinvestigated one of the most fundamental problems in structure light depth sensing field: correspondence retrieval of features between patterns and images. We formulate the global optimum correspondence retrieval by maximizing a conditional probability of correspondence given observed features, which is depicted by a Bayesian network. Different from traditional “code-only” based correspondence retrieval methods, the proposed Bayesian network based method exploits the positional correlations of correspondences of neighboring features, namely, the correspondences of poorly detected features are estimated with the aid of the correspondences of well detected features. The method performs especially well on challenging scenes with rich depth variations, abrupt depth changes, edges, etc. Experiments show that the proposed method increase the correspondence accuracy by about 40% on challenging scenes, compared with traditional “code-only” based correspondence retrieval methods.

This article introduces a new contrast enhancement algorithm of tone-preserving entropy maximization. Its design objective is to present the maximal amount of information content in the enhanced image, or being optimal in an information theoretical sense, while preventing the loss of tone continuity. The resulting optimization problem can be graph theoretically modeled as the construction of the $K$ -edges maximum-weight path, and it can be solved efficiently by dynamic programming. Moreover, the proposed algorithm is made more effective by being combined with a preprocess of image restoration that aims to correct quantization errors caused by the analog-to-digital conversion of image signals. Empirical evidences are provided to demonstrate the superior visual quality obtained by the new image enhancement algorithm.

In the field of affective computing (AC), coarse-grained AC has been developed and widely applied in many fields. Electroencephalogram (EEG) signals contain abundant emotional information. However, it is difficult to develop fine-grained AC due to the lack of fine-grained labeling data and suitable visualization methods for EEG data with fine labels. To achieve a fine mapping of EEG data directly to facial images, we propose a conditional generative adversarial network (cGAN) to establish the relationship between EEG data associated with emotions, a coarse label, and a facial expression image in this study. In addition, a corresponding training strategy is also proposed to realize the fine-grained estimation and visualization of EEG-based emotion. The experiments prove the reasonableness of the proposed method for the generation of fine-grained facial expressions. The image entropy of the generated image indicates that the proposed method can provide a satisfactory visualization of fine-grained facial expressions.

Fine-grained image recognition has been a hot research topic in computer vision due to its various applications. The-state-of-the-art is the part/region-based approaches that first localize discriminative parts/regions, and then learn their fine-grained features. However, these approaches have some inherent drawbacks: (1) the discriminative feature representation of an object is prone to be disturbed by complicated background; (2) it is unreasonable and inflexible to fix the number of salient parts, because the intended parts may be unavailable under certain circumstances due to occlusion or incompleteness, and (3) the spatial correlation among different salient parts has not been thoroughly exploited (if not completely neglected). To overcome these drawbacks, in this paper we propose a new, simple yet robust method by building part sequence model on the attended object region. Concretely, we first try to alleviate the background effect by using a region attention mechanism to generate the attended region from the original image. Then, instead of localizing different salient parts and extracting their features separately, we learn the part representation implicitly by applying a mapping function on the serialized features of the object. Finally, we combine the region attending network and the part sequence learning network into a unified framework that can be trained end-to-end with only image-level labels. Our extensive experiments on three fine-grained benchmarks show that the proposed method achieves the state of the art performance.

In this paper, we propose a deep propagation based image matting framework by introducing deep learning into learning an alpha matte propagation principal. Our deep learning architecture is a concatenation of a deep feature extraction module, an affinity learning module and a matte propagation module. These three modules are all differentiable and can be optimized jointly via an end-to-end training process. Our framework results in a semantic-level pairwise similarity of pixels for propagation by learning deep image representations adapted to matte propagation. It combines the power of deep learning and matte propagation and can therefore surpass prior state-of-the-art matting techniques in terms of both accuracy and training complexity, as validated by our experimental results from 243K images created based on two benchmark matting databases.

We consider the scene text recognition problem under the attention-based encoder-decoder framework, which is the state of the art. The existing methods usually employ a frame-wise maximal likelihood loss to optimize the models. When we train the model, the misalignment between the ground truth strings and the attention's output sequences of probability distribution, which is caused by missing or superfluous characters, will confuse and mislead the training process, and consequently make the training costly and degrade the recognition accuracy. To handle this problem, we propose a novel method called edit probability (EP) for scene text recognition. EP tries to effectively estimate the probability of generating a string from the output sequence of probability distribution conditioned on the input image, while considering the possible occurrences of missing/superfluous characters. The advantage lies in that the training process can focus on the missing, superfluous and unrecognized characters, and thus the impact of the misalignment problem can be alleviated or even overcome. We conduct extensive experiments on standard benchmarks, including the IIIT-5K, Street View Text and ICDAR datasets. Experimental results show that the EP can substantially boost scene text recognition performance.

Conventional sparse representation (SR)-based direction-of-arrival (DOA) estimation algorithms suffer from high computational complexity. To be specific, a wide angular range and a large-scale array will enlarge the scale of the spatial observation matrix, which results in huge computation cost for DOA estimation. In this letter, a new efficient DOA estimation algorithm based on the separable sparse representation (SSR-DOA for short) is derived, in which a separable structure for spatial observation matrix is introduced to reduce the complexity. Besides, a dual-sparsity strategy is engaged to make the algorithm tractable. Experimental results show that high resolution performance can be obtained efficiently by the proposed algorithm.

A single-shot structured light method is widely used to acquire dense and accurate depth maps for dynamic scenes. In this paper, we propose a color sequence coded fringe depth sensing method. To overcome the phase unwrapping problem encountered in phase-based methods, the color-coded sequence information is embedded into the phase information. We adopt the color-encoded De Bruijn sequence to denote the period of the phase information and assign the sequence into two channels of the pattern, while the third channel is used to code the phase information. Benefiting from this coding strategy, the phase information distributed in multiple channels can improve the quality of the phase intensity by channel overlay, which results in precise phase estimation. Meanwhile, the wrapped phase period assists the sequence decoding to obtain a precise period order. To evaluate the performance of the proposed method, an experimental platform is established. Quantitative and qualitative experiments demonstrate that the proposed method generates a higher precision depth, as compared to a Kinect and larger resolution ToF (Time of Flight) camera.

Due to its low complexity and acceptable accuracy, phase retrieval technique has been proposed as an alternative to solve the classic optical surface measurement task. However, to capture the overall wave field, phase retrieval based optical surface measurement (PROSM) system has to moderate the CCD position during the multiple-sampling procedure. The mechanical modules of CCD movement may bring about unexpectable deviation to the final results. To overcome this drawback, we propose a new PROSM method based on spatial light modulator (SLM). The mechanical CCD movement can be replaced by an electrical moderation of SLM patterns; thus the deviation can be significantly suppressed in the new PROSM method. In addition, to further improve the performance, we propose a new iterative threshold phase retrieval algorithm with sparsity-constraint to effectively reconstruct the phase of wave field. Experimental results show that the new method provides a more simple and robust solution for the optical surface measurement than the traditional techniques and achieves higher accuracy.

Resolution upconversion of a degraded image is an ill-posed inverse problem that is even harder than video superresolution due to lack of redundant observations from reference frames. To overcome this difficulty an adaptive 2D piecewise autoregressive (PAR) model is used to strengthen the constraints on the solution of the inverse problem. The PAR model can be fit to local image waveforms by adjusting its parameters. But estimating the model parameters needs the knowledge of the very original high-resolution pixels to be estimated by the model. We resolve this chicken-and-egg dilemma by adaptive nonlinear least-squares joint estimation of both model parameters and original pixels. This non-linear estimation problem is solved by the method of structured total least-squares, constrained by the degradation function (e.g., the point spread function of a camera plus noises) that forms the observed low-resolution image. As such, this work offers a unified general framework for joint upsampling, deconvolution, and denoising. Moreover, the upsampling can be carried out at an arbitrary scale rather than power of two. Experiments show that the proposed NEARU technique outperforms current methods in both PSNR and subjective visual quality, and its advantage becomes greater for larger scaling factors.

We develop a novel psychovisually motivated edge-based low-bit-rate image codec. It offers a compact description of scale-invariant second-order statistics of natural images, the preservation of which is crucial to the perceptual quality of coded images. Although being edge based, the codec does not explicitly code the edge geometry. To save bits on edge descriptions, a background layer of the image is first coded and transmitted, from which the decoder estimates the trajectories of significant edges. The edge regions are then refined by a residual coding technique based on edge dilation and sequential scanning in the edge direction. Experimental results show that the new image coding technique outperforms the existing ones in both objective and perceptual quality, particularly at low bit rates.

Although the current imaging sensors can achieve 12 or higher precision, the current display devices and the commonly used digital image formats are still only 8 bits. This mismatch causes significant waste of the sensor precision and loss of information when storing and displaying the images. For better usage of the precision-budget, tone mapping operators have to be used to map the high-precision data into low-precision digital images adaptively. In this paper, the classic histogram equalization tone mapping operator is reexamined in the sense of optimization. We point out that the traditional histogram equalization technique and its variants are fundamentally improper by suffering from local optimum problems. To overcome this drawback, we remodel the histogram equalization tone mapping task based on graphic theory which achieves the global optimal solutions. Another advantage of the graphic-based modeling is that the tone-continuity is also modeled as a vital constraint in our approach which suppress the annoying boundary artifacts of the traditional approaches. In addition, we propose a novel dynamic programming technique to solve the histogram equalization problem in real time. Experimental results shows that the proposed tone-preserved global optimal histogram equalization technique outperforms the traditional approaches by exhibiting more subtle details in the foreground while preserving the smoothness of the background.

Compressive sensing-based synthetic aperture radar (SAR) imaging has shown its superior capability in high-resolution image formation. However, most of those works focus on the scenes that can be sparsely represented in fixed spaces. When dealing with complicated scenes, these fixed spaces lack adaptivity in characterizing varied image contents. To solve this problem, a new compressive sensing-based radar imaging approach with adaptive sparse representation is proposed. Specifically, an autoregressive model is introduced to adaptively exploit the structural sparsity of an image. In addition, similarity among pixels is integrated into the autoregressive model to further promote the capability and thus an adaptive sparse representation facilitated by a weighted autoregressive model is derived. Since the weighted autoregressive model is inherently determined by the unknown image, we propose a joint optimization scheme by iterative SAR imaging and updating of the weighted autoregressive model to solve this problem. Eventually, experimental results demonstrated the validity and generality of the proposed approach.

A practical mobile-robot-based automatic transport service system is presented and discussed in this paper. A wheeled mobile robot, capable of two-directional movement, is developed to transport and feed materials to an industrial production line. For path guidance, an electromagnetic approach is investigated in detail. A conductor line laid on or under the ground defines the path, through which runs a square-wave current signal of a certain frequency, to produce an alternating magnetic field. Specific sensors are designed to measure this field and provide a path guidance signal for the robot. For path control, the proportional plus derivative (PD) algorithm is adopted to control the robot's movement along the path line. Multi-mode control strategies are developed to cover different path cases. The active turn control scheme is used for turning path modes. Finally, some practical testing results are presented.

A s an active and dynamic security-defense technique, intrusion detection can detect the interior and exterior attacks, and it plays an important role in assuring the network security. Based on immune recognition algorithm, a Radial Basis Function (RBF) neural network learning algorithm was studied. In this algorithm, the input data is regarded as antigens and antibodies are regarded as the hidden layer centers, the weights of the output layer are determined by adopting the Recursive Least Square method, which can improve convergence speed and precision of the RBF neural network, using Snort to establish innate antibody and using negative selection algorithm to generate detectors. This algorithm was applied to Intrusion Detection Systems. Theory and experiment show that this algorithm has better ability in intrusion detection, and can be used to improve the efficiency of intrusion detection, and reduce the false alarm rate.

With a single shot via 2D sensors Coded Aperture Snapshot Spectral Imaging (CASSI) system can restore 3D spatial-spectral images by solving a compressive sensing problem. But the shortage of measurements may lead to unfaithful reconstruction of 3D images. In addition due to the usage of random code aperture half of the energy is blocked which makes the imaging task more challengeable. In this paper we propose a double-channel spectral imaging system. The new framework adopts a beam splitter incorporated with two original CASSI systems employing a set of novel complementary code patterns. Complementary CS measurements are acquired in this framework. Furthermore to enhance the intensity of captured light we propose a more effective alternative solution based on digital-micromirror-device (DMD). All the incident light energy is captured by the sensors in the DMD-based framework. Simulation result shows that the proposed method significantly improves the performance of compressive spectral imaging.

A psychovisual quality driven image codec exploiting the psychological and neurological process of visual perception is proposed in this paper. Recent findings in brain theory and neuroscience suggest that visual perception is a process of fitting brain's internal generative model to the outside retina stimuli. And the psychovisual quality is related to how accurately visual sensory data can be explained by the internal generative model. Therefore, the design criterion of our psychovisually tuned image compression system is to find a compact description of the optimal generative model from the input image on the encoding end, which is then used to regenerate the output image on the decoding end. By noting an important finding from empirical natural image statistics that natural images have scale invariant features in the pixels' high order statistics, the generative model can be efficiently compressed through model preserving spatial downsampling on the encoder. And the decoder can reverse the process with a model preserving upsampling module to generate the decoded image. The proposed system is fully standard complaint because the downsampled image can be compressed with any exiting codec (JPEG2000 in this work). The proposed algorithm is shown to systematically outperform JPEG2000 in a wide bit rate range in terms of both subjective and objective qualities.

We propose a novel edge-based low bit-rate image codec. Although being edge based, the codec does not explicitly code the edge geometry. To avoid spending high bit budget on edges, a background layer of the image is first coded and transmitted, from which the decoder estimates the trajectories of significant edges. The edge regions are then refined by a residual coding technique based on edge dilation and sequential scanning in the edge direction. Experimental results show that the new image coding technique outperforms existing ones in both objective and perceptual quality, particularly at low bit rates.

An edge-based region-of-interest (ROI) image coding technique is proposed for low bit-rate visual communication. An image is compactly coded into two semantic levels: background sketch and object textures. The background sketch is a down-sampled version of the input image. The decoder reconstructs the background first by adaptive interpolation and lets users identify the ROI, and then requests the ROI textures to be transmitted. A distinct advantage of the proposed ROI technique is a compact edge-based descriptor of natural object boundaries. The expensive ROI geometry computations are carried out at the encoder only on demand, keeping decoder complexity low to benefit wireless devices. The new system outperforms the dynamic ROI coding of JPEG 2000 in both visuality and PSNR. Furthermore, both background and texture coding can be made compliant with any existing compression standard.

Web document classification is an important technique of web mining. Web pages classification has been studied extensively since the Internet has become a huge database of information. The k-NN is a simple classification algorithm that is used to assign patterns of unknown classification to the class of the majority of its k nearest neighbors of known classification according to the distance measure, but a main drawback of the method is that each of the patterns of known classification is considered equally important in the assignment of the pattern to be classified. Fuzzy k-nearest neighbor (fuzzy k-NN) is improving algorithm of k-NN, which is applied successfully in structural data classification. This paper presents the web document classification based on fuzzy k-NN network, in the process of classification, TF/IDF (term frequency / inverse document frequency) is adopted for selecting features of document, to increase the accuracy and suit for real world, membership grade is used. Experimental results show that classification performance is better than both k-NN and support vector machine (SVM).

In this paper we propose a new denoising technique based on improved adaptive directional lifting wavelet transform (ADL). Using this method can separate noise from image signal distinctly by the extraordinary ability of ADL to represent the edges and textures. However, in the smooth regions of a noisy image, ADL is expensive and inaccurate. Therefore, we construct the ADL in an anti-noise way based on pixel pattern classification. Compared with the traditional ADL for image compression, there is no restriction of side-information, so the optimal strategies of direction determination and transformation can be selected by the judgment of different pixel patterns. Experimental results show that the proposed technique outperforms traditional wavelet and lifting scheme in both PSNR and visual quality, especially for the images with rich texture features such as remote sensing images.

We propose a novel directional lifting-based block-set coding algorithm in this paper. Directional lifting wavelet transform based on region partition is firstly used to remove the correlation along the edges and reduce the amount of high-pass subband energy, then a new multi-symbol arithmetic coding based on residual classified context is exploited to code directional information. A context-based block-set partitioning method is proposed to code the directional transformed coefficients. Experimental results show that the proposed algorithm for texture-rich images outperforms JPEG2000 by 1.6 dB. Especially, the new algorithm can obtain the PSNR gain which up to 0.4 dB, comparing with JPEG2000 (EBCOT) based on directional lifting transform.

Recent researches have shown that the adaptive directional lifting (ADL) can represent edges and textures in images effectively. This makes it possible to separate noise from image signal distinctly in image denoising. However, a key issue named orientation estimation for ADL becomes inefficient and error prone in the noised circumstance. The authors propose a robust adaptive directional lifting-based (RADL) wavelet transform for image denoising by constructing ADL in an anti-noise way. In our method, a simple model of pixel pattern classification is incorporated into orientation estimation module to strengthen the robustness of this algorithm. Moreover, instead of determining the transform strategy based on sub-blocks, RADL is performed on pixel-level to pursue better denoising results. Experimental results show that the proposed technique demonstrates both PSNR and visual quality improvement on images with rich textures.

Monkey language models are defined for Chinese Phrase Networks, and scale-free features of Chinese Phrase Networks are uncovered. It is pointed out that the ratio of average degree to the total number of nodes (k/N) is close to a constant. Simulation for the evolution of phrase networks indicates that one of the important reasons for power law distributions is the word selection frequency, which, when tuned aptly, can make the monkey language present similar statistic traits as that of natural languages. Power law tails emerge at large k, and the exponent is about 6. Comparison between monkey model and natural language shows that humans are able to use Chinese words resources in more effective and compact ways to express their intentions. All the results demonstrate an important fact that the least effort principle is the basis of Chinese Phrase Networks.