Search

We present our automated real-time socket inspection system capable of detecting an assortment of defects including metallic and liquid staining, loose capacitors and pins, and other debris and foreign material (FM). Our test tools pick and place manufactured units into sockets for electrical testing. Any debris accumulated inside the test sockets will likely damage subsequent units until the defective socket is replaced. To quickly capture in-situ defects and mitigate further damage, we equipped each pick-and-place arm with a new vision system designed to fit within the existing tool. The tight footprint constraints required a highly compact imaging system which resulted in a variety of image artifacts, creating several unique challenges. Our inspection algorithm utilizes a variety of advanced computer vision and machine learning techniques to normalize images, extract and match features, register the images, suppress unwanted artifacts, and detect defects. The detected changes are then sent to a deep learning classifier to further filter between true defects and natural socket deterioration. The flagged socket images can be manually dispositioned by the user and the socket can be sent for repair or cleaning as needed. [ABSTRACT FROM AUTHOR]

Classification of 3D objects – the selection of a category in which each object belongs – is of great interest in the field of machine learning. Numerous researchers use deep neural networks to address this problem, altering the network architecture and representation of the 3D shape used as an input. To investigate the effectiveness of their approaches, we conduct an extensive survey of existing methods and identify common ideas by which we categorize them into a taxonomy. Second, we evaluate 11 selected classification networks on two 3D object datasets, extending the evaluation to a larger dataset on which most of the selected approaches have not been tested yet. For this, we provide a framework for converting shapes from common 3D mesh formats into formats native to each network, and for training and evaluating different classification approaches on this data. Despite being partially unable to reach the accuracies reported in the original papers, we compare the relative performance of the approaches as well as their performance when changing datasets as the only variable to provide valuable insights into performance on different kinds of data. We make our code available to simplify running training experiments with multiple neural networks with different prerequisites. [ABSTRACT FROM AUTHOR]

The wire bonding process is one of the most critical processes in semiconductor packaging. The electrical performance and reliability of IC chips must be evaluated by probe testing before wire bonding. In the probe test, the probe usually leaves traces on the surface of the pad. Large probe marks tend to cause a decrease in bond adhesion and so increase the possibility of ball bond lifting from the pads. Shear force is an important measure of the adhesion between the ball and pad. The prediction of the shear force helps to understand the bonding quality in advance. In this study, the six features of the probe marks were extracted by the automatic image recognition method. Using three machine learning techniques (logistic regression, support vector machine, and random forest) based on principal component analysis (PCA), the shear force is estimated based on six features before wire bonding. The results indicate that the proposed PCA-based random forest could identify bad chips with 97.92% accuracy before wire bonding. Unnecessary chips can be found early, and unnecessary wire bonding can be avoided, thereby saving process time and cost. The proposed method would improve the efficiency and quality yield of the semiconductor packaging process. [ABSTRACT FROM AUTHOR]

Spatial failure patterns in wafer defect maps are very useful for root cause analysis, which is essential for yield optimization in the manufacturing process. Especially, the scratch defect type is the most challenging pattern to recognize because the position, shape, size and curvature vary widely from one scratch to another. Discontinuity points within scratches also contribute to the low recognition rate, and such points are often hidden defective dies that become reliability threat. Previous studies on defect pattern identification show that the recognition rate for scratches is among the lowest in all patterns even if the overall accuracy is high. In this paper, we propose a novel scratch pattern recognition method. In this paper, image processing techniques are applied to recognize scratch patterns in wafer maps. The Hough transform is first employed to identify line segments, from which scratches can be reconstructed. In contrast to previous machine-learning based methods, there is no need to train a complicated prediction model so that the computation time is small. The method is validated by using wafer maps from six real products and public data. Experimental results show that the proposed method achieves high recall, precision and accuracy in all cases. [ABSTRACT FROM AUTHOR]

Multi-view learning has become a significant research topic in image processing, data mining and machine learning due to the proliferation of multi-view data. Considering the difficulty in obtaining labeled data in many real applications, we focus on the multi-view unsupervised feature selection problem. Most existing multi-view feature selection introduce an identical similarity matrix among different views, which cannot preserve the specific correlation between each single view. Also, some of these methods just consider either global or local structures. In this paper, we propose an embedding method, Adaptive Similarity Embedding for Unsupervised Multi-View Feature Selection (ASE-UMFS). This method reduces the high-dimensional data to the low dimensions and unifies different views to a combination weight matrix. We also use parameters to constraint the similarity matrix for the local structure, where the regularization term is used to add a prior of uniform distribution; taking into account of the independence in projection matrix among different views, optimization of the similarity matrix is further improved. To confirm the effectiveness of ASE-UMFS, comparisons are made with benchmark algorithm on real-world data sets. The experimental results demonstrate that the proposed algorithm outperforms several state-of-the-art methods in multi-view learning. [ABSTRACT FROM AUTHOR]

Deep learning algorithms, especially convolutional neural networks (CNNs), have recently emerged as a dominant paradigm for high spatial resolution remote sensing (HRS) image recognition. A large amount of CNNs have already been successfully applied to various HRS recognition tasks, such as land-cover classification and scene classification. However, they are often modifications of the existing CNNs derived from natural image processing, in which the network architecture is inherited without consideration of the complexity and specificity of HRS images. In this article, the remote sensing deep neural network (RSNet) framework is proposed using an automatically search strategy to find the appropriate network architecture for HRS image recognition tasks. In RSNet, the hierarchical search space is first designed to include module- and transition-level spaces. The module-level space defines the basic structure block, where a series of lightweight operations as candidates, including depthwise separable convolutions, is proposed to ensure the efficiency. The transition-level space controls the spatial resolution transformations of the features. In the hierarchical search space, a gradient-based search strategy is used to find the appropriate architecture. In RSNet, the task-driven architecture training process can acquire the optimal model parameters of the switchable recognition module for HRS image recognition tasks. The experimental results obtained using four benchmark data sets for land-cover classification and scene classification tasks demonstrate that the searched RSNet can achieve a satisfactory accuracy with a high computational efficiency and, hence, provides an effective option for the processing of HRS imagery. [ABSTRACT FROM AUTHOR]

In lifelong machine learning (LML) systems, consecutive new tasks from changing circumstances are learned and added to the system. However, sufficiently labeled data are indispensable for extracting intertask relationships before transferring knowledge in classical supervised LML systems. Inadequate labels may deteriorate the performance due to the poor initial approximation. In order to extend the typical LML system, we propose a novel hierarchical lifelong learning algorithm (HLLA) consisting of two following layers: 1) the knowledge layer consisted of shared representations and integrated knowledge basis at the bottom and 2) parameterized hypothesis functions with features at the top. Unlabeled data is leveraged in HLLA for pretraining of the shared representations. We also have considered a selective inherited updating method to deal with intertask distribution shifting. Experiments show that our HLLA method outperforms many other recent LML algorithms, especially when dealing with higher dimensional, lower correlation, and fewer labeled data problems. [ABSTRACT FROM AUTHOR]

Visual odometry (VO) is one of the essential techniques in mobile robots field; an accurate VO system is of great significance for mobile robot simultaneous localization and mapping. As for traditional monocular VO systems, they work by presuming the monocular scale is 1 (scale = 1), or relying on ground truth (GT) to estimate scale. As a result, the traditional monocular VO systems estimate the pose state with big drift or cannot work on the image sequence without GT. Although some classical monocular VO systems have been proposed, they still have imperfect performance or even unable to work in some extreme scene conditions, such as scene is monotony without obvious texture information or camera large-scale displacement motion. As for learning-based VO system, it is realized by training deep neural networks in supervised or self-supervised manner to end-to-end estimate the pose state; however, the accuracy of pose estimation entirely depends on the ability of networks. Although the ability of networks can be improved by increasing the number of training data sets and optimizing the network structure, it is inevitable to encounter problems such as insufficient generalization ability and insufficient accuracy on rotational pose estimation. In this article, a monocular VO system named DL_Hybrid is proposed, which takes full advantage of DL networks used in image processing and geometric localization theory based on hybrid pose estimation methods. The DL_Hybrid VO system can estimate a six-DoF pose one-frame-by-one-frame and recover camera trajectory, and it can extract accurate key points from per-frame even in extreme scene condition, and it has good performance even in the extreme moving condition, such as camera rotation-only action or static action, also it can work well in the condition of camera large-scale displacement motion. The real scale is also accurately estimated without depending on GT, and the pose estimation method is designed based on hybrid 2d–2d and 3d–2d localization theory to make the DL_Hyrid VO system to estimate translational and rotational information with accuracy and robustness. Experimental results show that the proposed DL_Hybrid VO system has a better performance than traditional and learning-based VO systems. [ABSTRACT FROM AUTHOR]

High/very-high-resolution (HR/VHR) multitemporal images are important in remote sensing to monitor the dynamics of the Earth’s surface. Unsupervised object-based image analysis provides an effective solution to analyze such images. Image semantic segmentation assigns pixel labels from meaningful object groups and has been extensively studied in the context of single-image analysis, however not explored for multitemporal one. In this article, we propose to extend supervised semantic segmentation to the unsupervised joint semantic segmentation of multitemporal images. We propose a novel method that processes multitemporal images by separately feeding to a deep network comprising of trainable convolutional layers. The training process does not involve any external label, and segmentation labels are obtained from the argmax classification of the final layer. A novel loss function is used to detect object segments from individual images as well as establish a correspondence between distinct multitemporal segments. Multitemporal semantic labels and weights of the trainable layers are jointly optimized in iterations. We tested the method on three different HR/VHR data sets from Munich, Paris, and Trento, which shows the method to be effective. We further extended the proposed joint segmentation method for change detection (CD) and tested on a VHR multisensor data set from Trento. [ABSTRACT FROM AUTHOR]

With an increasing number of remote sensing images (RSIs), the automatic region of interest (ROI) extraction based on convolutional neural networks (CNNs) has attracted much interest in recent years. Although fully supervised CNN-based methods have shown superiority in the field of object extraction, pixelwise annotations are expensive and time-consuming. Moreover, due to the unstable distribution of ROIs in complex landscapes, the ratios of foreground and background areas are quite different in RSIs. Training CNNs with such unbalanced data sets lead to over-fitting and low accuracy. In this article, we propose a framework that combines multiview learning and attention mechanism (MLAM) to solve the above mentioned problems. First, we develop a CNN-based weakly supervised method with a weight-balanced loss function to solve the problems caused by an unbalanced data set. It also helps to generate imagewise saliency maps by computing the gradient maps with respect to the input images. Then, we design a multiview strategy to dramatically reduce the missing inspection. Finally, we design a feedback attention mechanism based on the stage neighbor binary pattern to further modify the extraction result. In summary, the proposed framework achieves pixelwise ROI extraction under imagewise annotations through data-driven ML and a knowledge-driven visual attention mechanism. We evaluate the performance of the MLAM framework on two challenging data sets with complex backgrounds. The experimental results indicate that the proposed framework can achieve better performance than other eight ROI extraction models for unbalanced remote sensing data sets. [ABSTRACT FROM AUTHOR]

This article proposes a new volcano seismic signal descriptor for improving the area under the receiver operating characteristic curve (AUC) in the classification of long-period (LP) and volcano-tectonic (VT) seismic events. It aims to describe a volcanic seismic event from a different and novel point of view that involves image processing techniques instead of classical seismic signal processing strategies, such as frequency or scale analysis. The proposed descriptor allows exploring the seismic signal space for obtaining the determination of the event patterns and, subsequently, the extraction of intensity-, shape-, and texture-based features into a numeric vectorial output for supplying a set of selected machine learning classifiers with different taxonomies. The descriptor was validated on a seismic signal database collected at the Cotopaxi volcano, containing a total of 637 events, including LP, VT, and other types of seismic events (e.g., rockfall or icequakes). An accuracy value of 96% was obtained in the determination of the event patterns using the signal database, while the values of 0.95 and 0.96 were obtained for the AUC when using a feedforward backpropagation artificial neural network classifier on two experimental data sets, containing feature vectors representing signal with and without event overlapping, respectively. The obtained results demonstrate that the proposed descriptor is capable of providing adequate seismic signal representations in a different feature space and that its output provides competitive results in the classification of volcanic seismic events. [ABSTRACT FROM AUTHOR]

Precision positioning is a very important task for automatic assembly and inspection in the manufacturing process. The conventional image processing for image alignment has relied on template matching, which is computationally intensive for objects in arbitrary locations and orientations. In this article, we propose the deep neural network regressors for fast and accurate image alignment. They are especially applied to the positioning of printed circuit boards (PCBs). The simple multilayer perceptron (MLP), the convolutional neural network (CNN), and the CNN models incorporated with support vector regression (SVR) are proposed and evaluated for the PCB positioning task. The proposed deep neural networks require only one single reference sample with a manually marked template window. All training images and the ground-truth geometric parameters are automatically generated for the model training. The effect of illumination changes and the strategies to cope with lighting variations are analyzed and proposed for robust positioning. Experimental results indicate that the proposed regressors can achieve a subpixel accuracy in translation and yield a rotation error less than 1° with 1-ms evaluation time for PCB positioning. [ABSTRACT FROM AUTHOR]

The family of image visibility graphs (IVG/IHVGs) have been recently introduced as simple algorithms by which scalar fields can be mapped into graphs. Here we explore the usefulness of such\an operator in the scenario of image processing and image classification. We demonstrate that the link architecture of the image visibility graphs encapsulates relevant information on the structure of the images and we explore their potential as image filters. We introduce several graph features, including the novel concept of Visibility Patches, and show through several examples that these features are highly informative, computationally efficient and universally applicable for general pattern recognition and image classification tasks. [ABSTRACT FROM AUTHOR]

Currently, deep neural networks (DNNs) are an important method for handling hyperspectral image (HSI) classification because of their good performance in image processing. However, DNNs’ performance depends on a massive number of training data and hyperparameters that are carefully fine-tuned, which results in structural complexity and a time-consuming process. Deep forest is a novel deep learning method that does not need much training data and has a simple structure. In this paper, we first design a deep forest for spectral-based HSI classification and then propose an improved deep forest algorithm, named deep multigrained cascade forest (dgcForest), for spatial-based HSI classification. On the one hand, the cascade forest in dgcForest is used in multigrained scanning, which enhances the performance; on the other hand, a pooling layer is added after the multigrained scanning to reduce the output dimensions. To demonstrate that our proposed algorithm presents a good performance in HSI classification, we analyze the hyperparameters of deep forest and dgcForest and compare them with other methods on the biased and unbiased data sets, which illustrates that our method is superior to other state-of-the-art deep learning methods. [ABSTRACT FROM AUTHOR]

Hyperspectral image (HSI) classification has become a hot topic in the field of remote sensing. In general, the complex characteristics of hyperspectral data make the accurate classification of such data challenging for traditional machine learning methods. In addition, hyperspectral imaging often deals with an inherently nonlinear relation between the captured spectral information and the corresponding materials. In recent years, deep learning has been recognized as a powerful feature-extraction tool to effectively address nonlinear problems and widely used in a number of image processing tasks. Motivated by those successful applications, deep learning has also been introduced to classify HSIs and demonstrated good performance. This survey paper presents a systematic review of deep learning-based HSI classification literatures and compares several strategies for this topic. Specifically, we first summarize the main challenges of HSI classification which cannot be effectively overcome by traditional machine learning methods, and also introduce the advantages of deep learning to handle these problems. Then, we build a framework that divides the corresponding works into spectral-feature networks, spatial-feature networks, and spectral–spatial-feature networks to systematically review the recent achievements in deep learning-based HSI classification. In addition, considering the fact that available training samples in the remote sensing field are usually very limited and training deep networks require a large number of samples, we include some strategies to improve classification performance, which can provide some guidelines for future studies on this topic. Finally, several representative deep learning-based classification methods are conducted on real HSIs in our experiments. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNNs) have been applied to visual tracking with demonstrated success in recent years. Most CNN-based trackers utilize hierarchical features extracted from a certain layer to represent the target. However, features from a certain layer are not always effective for distinguishing the target object from the backgrounds especially in the presence of complicated interfering factors (e.g., heavy occlusion, background clutter, illumination variation, and shape deformation). In this work, we propose a CNN-based tracking algorithm which hedges deep features from different CNN layers to better distinguish target objects and background clutters. Correlation filters are applied to feature maps of each CNN layer to construct a weak tracker, and all weak trackers are hedged into a strong one. For robust visual tracking, we propose a hedge method to adaptively determine weights of weak classifiers by considering both the difference between the historical as well as instantaneous performance, and the difference among all weak trackers over time. In addition, we design a Siamese network to define the loss of each weak tracker for the proposed hedge method. Extensive experiments on large benchmark datasets demonstrate the effectiveness of the proposed algorithm against the state-of-the-art tracking methods. [ABSTRACT FROM AUTHOR]

Deep learning methods, especially CNNs, have achieved many promising results in a wide range of computer vision applications. However, few studies focused on designing suitable deep learning methods for parking space status inference. As we have known, it is challenging to detect parking spaces in an outdoor environment due to dynamic lighting variations, weather changes, and perspective distortion. By off-the-shelf CNNs, lighting variations might be handled well. However, to realize a practical and robust inference system, we also need to address troublesome problems, such as parking displacements, non-unified car sizes, inter-object occlusion, and perspective distortion. These problems may become even challenging if also considering the difference of space sizes. To overcome the problems, we proposed a custom–tailored deep convolutional and contrastive network with three contributions. First, we introduced a Siamese architecture to learn the contrastive and robust feature descriptor. This helps to reduce the effects owing to the variety of inter-object occlusion. Second, we integrated a convolutional Spatial Transformer Network (STN) to adaptively transform a 3-space input patch according to vehicle sizes and parking displacement. STN also helps to overcome the perspective distortion problem. Third, a multi-task loss function was designed to train the network by simultaneously considering the accuracy of inferring the status of the target space and the semantic smoothness of high-level features. Thereby, the errors caused by inter-object occlusion could be alleviated. To verify the proposed network, we visualized the learned features and analyzed their functionality. Experiments and evaluations have shown the robustness of our system in parking status inference. The real-time system currently running in public parking lots also demonstrates the effectiveness of the proposed deep network. [ABSTRACT FROM AUTHOR]

In this paper, we present a comparative evaluation of various multi-person tracking-by-detection approaches on public data sets. This paper investigates five popular trackers coupled with six relevant visual people detectors evaluated on seven public data sets. The evaluation emphasizes on exhibited performance variation depending on tracker-detector choices. Our experimental results show that the overall performance depends on how challenging the data set is, the performance of the detector on the specific data set, and the tracker-detector combination. Some trackers are more sensitive to the choice of a detector and some detectors to the choice of a tracker than others. Based on our results, two of the trackers demonstrate the best performances consistently across different data sets, whereas the best performing detectors vary per data set. This underscores the need for careful application context specific evaluation when choosing a detector. [ABSTRACT FROM AUTHOR]

Dynamic scene classification has been extensively studied in computer vision due to its widespread applications. The key to dynamic scene classification lies in jointly characterizing spatial appearance and temporal dynamics to achieve informative representation, which remains an outstanding task in the literature. In this paper, we propose a unified framework to extract spatial and temporal features for dynamic scene representation. More specifically, we deploy two variants of deep convolutional neural networks to encode spatial appearance and short-term dynamics into short-term deep features (STDF). Based on STDF, we propose using the autoregressive moving average model to extract long-term frequency features (LTFF). By combining STDF and LTFF, we establish the long–short-term feature (LSTF) representations of dynamic scenes. The LSTF characterizes both spatial and temporal patterns of dynamic scenes for comprehensive and information representation that enables more accurate classification. Extensive experiments on three-dynamic scene classification benchmarks have shown that the proposed LSTF achieves high performance and substantially surpasses the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Achieving good text detection and recognition results for multi-script-oriented images is a challenging task. First, we explore bit plane slicing in order to utilize the advantage of the most significant bit information to identify text components. A new iterative nearest neighbor symmetry is then proposed based on shapes of convex and concave deficiencies of text components in bit planes to identify candidate planes. Further, we introduce a new concept called mutual nearest neighbor pair components based on gradient direction to identify representative pairs of texts in each candidate bit plane. The representative pairs are used to restore words with the help of edge image of the input one, which results in text detection results (words). Second, we propose a new idea by fixing window for character components of arbitrary oriented words based on angular relationship between sub-bands and a fused band. For each window, we extract features in contourlet wavelet domain to detect characters with the help of an SVM classifier. Further, we propose to explore HMM for recognizing characters and words of any orientation using the same feature vector. The proposed method is evaluated on standard databases such as ICDAR, YVT video, ICDAR, SVT, MSRA scene data, ICDAR born digital data, and multi-lingual data to show its superiority to the state of the art methods. [ABSTRACT FROM AUTHOR]

A novel scheme of edge detection based on the physical law of diffusion is presented in this paper. Though the most current studies are using data based methods such as deep neural networks, these methods on machine learning need big data of labeled ground truth as well as a large amount of resources for training. On the other hand, the widely used traditional methods are based on the gradient of the grayscale or color of images with using different sorts of mathematical tools to accomplish the mission. Instead of treating the outline of an object in an image as a kind of gradient of grayscale or color, our scheme deals with the edge detection as a character of an energy diffusing in the space of media such as charge-coupled device. By using the characteristic function of diffusion, the information of the energy will be extracted. The scheme preserves the structural information of images very well. Because it comes from the inhere law of images’ physical property, it has a unified mathematical framework for images’ edge detection under different conditions, for example, multiscales, diferent light conditions, and so on. Moreover, it has low computational complexity. [ABSTRACT FROM AUTHOR]

Recognizing objects from simultaneously sensed photometric (RGB) and depth channels is a fundamental yet practical problem in many machine vision applications, such as robot grasping and autonomous driving. In this paper, we address this problem by developing a cross-modal attentional context (CMAC) learning framework, which enables the full exploitation of the context information from both RGB and depth data. Compared to existing RGB-D object detection frameworks, our approach has several appealing properties. First, it consists of an attention-based global context model for exploiting adaptive contextual information and incorporating this information into a region-based CNN (e.g., fast RCNN) framework to achieve improved object detection performance. Second, our CMAC framework further contains a fine-grained object part attention module to harness multiple discriminative object parts inside each possible object region for superior local feature representation. While greatly improving the accuracy of RGB-D object detection, the effective cross-modal information fusion as well as attentional context modeling in our proposed model provide an interpretable visualization scheme. Experimental results demonstrate that the proposed method significantly improves upon the state of the art on all public benchmarks. [ABSTRACT FROM AUTHOR]

Polarimetric synthetic aperture radar (PolSAR) image classification plays an important role in remote sensing image processing. In recent years, stacked auto-encoder (SAE) has obtained a series of excellent results in PolSAR image classification. The recently proposed projective dictionary pair learning (DPL) model takes both accuracy and time consumption into consideration, and another recently proposed semicoupled dictionary learning (SCDL) model gives a new way to fit different features. Based on the SAE, DPL, and SCDL models, we propose a novel semicoupled projective DPL method with SAE (SAE-SDPL) for PolSAR image classification. Our method can get the classification result efficiently and correctly and meanwhile giving a new method to fit different features. In this paper, three PolSAR images are used to test the performance of SAE-SDPL. Compared with some state-of-the-art methods, our method obtains excellent results in PolSAR image classification. [ABSTRACT FROM AUTHOR]

In this paper, we propose a hierarchical polarization feature generation using a self-organizing codebook to realize unsupervised land classification that fully utilizes the detailed polarization information contained in high-resolution polarimetric synthetic aperture radar (PolSAR) data. PolSAR has reached a decimeter-level high resolution. In general, conventional methods lower the resolution of the PolSAR data to 10–20 m in the real-space distance to classify observation regions into land classes such as farm, forest, and town. However, lowering resolution prevents us from discovering new land classes potentially enabled by the resolution enhancement. The hierarchical method we propose here not only classifies observation regions successfully into land classes such as farm, forest, and town that humans can naturally distinguish but also discovers new land subclasses findable only in high-resolution PolSAR data. We explain these two types of our achievements (classification/discovery) through experimental results for Japan Aerospace Exploration Agency’s polarimetric and interferometric airborne SAR-L2 data having decimeter resolution. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have recently shown excellent performance on the task of fine-grained vehicle classification, where the motivation is to identify the fine-grained categories of the given vehicles. Generally speaking, the main motivation of the conventional back-propagation algorithm is to optimize the loss function. The algorithm itself does not guarantee if the extracted features are discriminative for the task of classification. Intuitively, if we can learn more discriminative features with a relatively small number of feature maps, the generalization ability of the CNNs will be significantly improved. Therefore, we propose a channel max pooling (CMP) scheme, where a new layer is inserted between the fully connected layers and the convolutional layers. The proposed CMP scheme divides the feature maps into to several sub-groups. Then, it compresses the feature maps within each sub-group into a new one. The compression is carried out by selecting the maximum value among the same locations from different feature maps. Moreover, the proposed CMP layer has the advantage that it can reduce the number of parameters via reducing the number of channels in the CNNs. Experimental results on two fine-grained vehicle datasets demonstrate that the CMP modified CNNs can improve the classification accuracies on the task of fine-grained vehicle classification while a massive amount of parameters are reduced. Moreover, it has competitive performance when comparing with the-state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Person re-identification (re-id) is a critical problem in video analytics applications such as security and surveillance. The public release of several datasets and code for vision algorithms has facilitated rapid progress in this area over the last few years. However, directly comparing re-id algorithms reported in the literature has become difficult since a wide variety of features, experimental protocols, and evaluation metrics are employed. In order to address this need, we present an extensive review and performance evaluation of single- and multi-shot re-id algorithms. The experimental protocol incorporates the most recent advances in both feature extraction and metric learning. To ensure a fair comparison, all of the approaches were implemented using a unified code library that includes 11 feature extraction algorithms and 22 metric learning and ranking techniques. All approaches were evaluated using a new large-scale dataset that closely mimics a real-world problem setting, in addition to 16 other publicly available datasets: VIPeR, GRID, CAVIAR, DukeMTMC4ReID, 3DPeS, PRID, V47, WARD, SAIVT-SoftBio, CUHK01, CHUK02, CUHK03, RAiD, iLIDSVID, HDA+, and Market1501. The evaluation codebase and results will be made publicly available for community use. [ABSTRACT FROM AUTHOR]

In this paper, we propose a spectral-spatial feature extraction and classification framework based on an artificial neuron network in the context of hyperspectral imagery. With limited labeled samples, only spectral information is exploited for training and spatial context is integrated posteriorly at the testing stage. Taking advantage of recent advances in face recognition, a joint supervision symbol that combines softmax loss and center loss is adopted to train the proposed network, by which intraclass features are gathered while interclass variations are enlarged. Based on the learned architecture, the extracted spectrum-based features are classified by a center classifier. Moreover, to fuse the spectral and spatial information, an adaptive spectral-spatial center classifier is developed, where multiscale neighborhoods are considered simultaneously, and the final label is determined using an adaptive voting strategy. Finally, experimental results on three well-known data sets validate the effectiveness of the proposed methods compared with the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Remote sensing image scene classification is a fundamental problem, which aims to label an image with a specific semantic category automatically. Recently, deep learning methods have achieved competitive performance for remote sensing image scene classification, especially the methods based on a convolutional neural network (CNN). However, most of the existing CNN methods only use feature vectors of the last fully connected layer. They give more importance to global information and ignore local information of images. It is common that some images belong to different categories, although they own similar global features. The reason is that the category of an image may be highly related to local features, other than the global feature. To address this problem, a method based on rearranged local features is proposed in this paper. First, outputs of the last convolutional layer and the last fully connected layer are employed to depict the local and global information, respectively. After that, the remote sensing images are clustered to several collections using their global features. For each collection, local features of an image are rearranged according to their similarities with local features of the cluster center. In addition, a fusion strategy is proposed to combine global and local features for enhancing the image representation. The proposed method surpasses the state of the arts on four public and challenging data sets: UC-Merced, WHU-RS19, Sydney, and AID. [ABSTRACT FROM AUTHOR]

With the rapidly increasing demand for security and E-health applications, device-free human detection has attracted interest because it does not require a wearable device or camera setup. This paper proposes a deep-learning-based approach that monitors wireless signals to learn three human modes, i.e., absence, working, and sleeping, in realistic indoor environments. This paper integrates the amplitude and phase of channel state information to propose a hybrid complex feature; this facilitates robust and efficient human detection even with fewer data samples. Experiments conducted in two unmodified WiFi networks demonstrate the effectiveness of the proposed algorithms. Four machine-learning algorithms provide satisfactory performance with sufficient data, and deep neural networks perform the best. Results show that by using 6% training samples, the proposed hybrid feature still achieves 93% accuracy and can even outperform three typical machine learning algorithms that use full training samples. Moreover, the proposed feature significantly improves detection accuracy by 11.62%–27.76% than traditional amplitude feature with fewer training samples. [ABSTRACT FROM AUTHOR]

Linear discriminant analysis (LDA) is a very popular supervised feature extraction method and has been extended to different variants. However, classical LDA has the following problems: 1) The obtained discriminant projection does not have good interpretability for features; 2) LDA is sensitive to noise; and 3) LDA is sensitive to the selection of number of projection directions. In this paper, a novel feature extraction method called robust sparse linear discriminant analysis (RSLDA) is proposed to solve the above problems. Specifically, RSLDA adaptively selects the most discriminative features for discriminant analysis by introducing the $l_{2,1}$ norm. An orthogonal matrix and a sparse matrix are also simultaneously introduced to guarantee that the extracted features can hold the main energy of the original data and enhance the robustness to noise, and thus RSLDA has the potential to perform better than other discriminant methods. Extensive experiments on six databases demonstrate that the proposed method achieves the competitive performance compared with other state-of-the-art feature extraction methods. Moreover, the proposed method is robust to the noisy data. [ABSTRACT FROM AUTHOR]

High-dimensional data contain not only redundancy but also noises produced by the sensors. These noises are usually non-Gaussian distributed. The metrics based on Euclidean distance are not suitable for these situations in general. In order to select the useful features and combat the adverse effects of the noises simultaneously, a robust sparse subspace learning method in unsupervised scenario is proposed in this paper based on the maximum correntropy criterion that shows strong robustness against outliers. Furthermore, an iterative strategy based on half quadratic and an accelerated block coordinate update is proposed. The convergence analysis of the proposed method is also carried out to ensure the convergence to a reliable solution. Extensive experiments are conducted on real-world data sets to show that the new method can filter out the outliers and outperform several state-of-the-art unsupervised feature selection methods. [ABSTRACT FROM AUTHOR]

In this paper, we propose a Weighted Local sparse representation based Depth Image Super-Resolution (WLDISR) schemes aiming at improving the Virtual View Image (VVI) quality of 3D video system. Different from color images, depth images are mainly used to provide geometrical information in synthesizing VVI. Due to the view synthesis characteristics difference between textural structures and smooth regions of depth images, we divide the depth images into edge and smooth patches and learn two local dictionaries, respectively. Meanwhile, the weight term is derived and incorporated explicitly in the cost function to denote different importance of edge structures and smooth regions to the VVI quality. Then, local sparse representation and weighted sparse representation are jointly used in both dictionary learning and reconstruction phases in depth image super-resolution. Based on different optimizations on learning and reconstruction modules, three WLDISR schemes, WLDISR-D, WLDISR-R, and WLDISR-ALL, are proposed. Experimental results on 3D sequences demonstrate that the proposed WLDISR-D, WLDISR-R, and WLDISR-ALL schemes can achieve more than 1.9-, 2.03-, and 2.16-dB gains on average, respectively, in terms of the VVIs’ quality, as compared with the state-of-the-art schemes. In addition, the visual quality of VVIs is also improved. [ABSTRACT FROM AUTHOR]

Scene classification of remote sensing images has drawn great attention because of its wide applications. In this paper, with the guidance of the human visual system (HVS), we explore the attention mechanism and propose a novel end-to-end attention recurrent convolutional network (ARCNet) for scene classification. It can learn to focus selectively on some key regions or locations and just process them at high-level features, thereby discarding the noncritical information and promoting the classification performance. The contributions of this paper are threefold. First, we design a novel recurrent attention structure to squeeze high-level semantic and spatial features into several simplex vectors for the reduction of learning parameters. Second, an end-to-end network named ARCNet is proposed to adaptively select a series of attention regions and then to generate powerful predictions by learning to process them sequentially. Third, we construct a new data set named OPTIMAL-31, which contains more categories than popular data sets and gives researchers an extra platform to validate their algorithms. The experimental results demonstrate that our model makes great promotion in comparison with the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) exhibit good performance in image processing tasks, pointing themselves as the current state-of-the-art of deep learning methods. However, the intrinsic complexity of remotely sensed hyperspectral images still limits the performance of many CNN models. The high dimensionality of the HSI data, together with the underlying redundancy and noise, often makes the standard CNN approaches unable to generalize discriminative spectral–spatial features. Moreover, deeper CNN architectures also find challenges when additional layers are added, which hampers the network convergence and produces low classification accuracies. In order to mitigate these issues, this paper presents a new deep CNN architecture specially designed for the HSI data. Our new model pursues to improve the spectral–spatial features uncovered by the convolutional filters of the network. Specifically, the proposed residual-based approach gradually increases the feature map dimension at all convolutional layers, grouped in pyramidal bottleneck residual blocks, in order to involve more locations as the network depth increases while balancing the workload among all units, preserving the time complexity per layer. It can be seen as a pyramid, where the deeper the blocks, the more feature maps can be extracted. Therefore, the diversity of high-level spectral–spatial attributes can be gradually increased across layers to enhance the performance of the proposed network with the HSI data. Our experiments, conducted using four well-known HSI data sets and 10 different classification techniques, reveal that our newly developed HSI pyramidal residual model is able to provide competitive advantages (in terms of both classification accuracy and computational time) over the state-of-the-art HSI classification methods [ABSTRACT FROM AUTHOR]

Vasculature is known to be of key biological significance, especially in the study of tumors. As such, considerable effort has been focused on the automated segmentation of vasculature in medical and pre-clinical images. The majority of vascular segmentation methods focus on bloodpool labeling methods; however, particularly, in the study of tumors, it is of particular interest to be able to visualize both the perfused and the non-perfused vasculature. Imaging vasculature by highlighting the endothelium provides a way to separate the morphology of vasculature from the potentially confounding factor of perfusion. Here, we present a method for the segmentation of tumor vasculature in 3D fluorescence microscopic images using signals from the endothelial and surrounding cells. We show that our method can provide complete and semantically meaningful segmentations of complex vasculature using a supervoxel-Markov random field approach. We show that in terms of extracting meaningful segmentations of the vasculature, our method outperforms both state-of-the-art method, specific to these data, as well as more classical vasculature segmentation methods. [ABSTRACT FROM AUTHOR]

Example learning-based single image super-resolution (SR) is a promising method for reconstructing a high-resolution (HR) image from a single-input low-resolution (LR) image. Lots of popular SR approaches are more likely either time-or space-intensive, which limit their practical applications. Hence, some research has focused on a subspace view and delivered state-of-the-art results. In this paper, we utilize an effective way with mixture prior models to transform the large nonlinear feature space of LR images into a group of linear subspaces in the training phase. In particular, we first partition image patches into several groups by a novel selective patch processing method based on difference curvature of LR patches, and then learning the mixture prior models in each group. Moreover, different prior distributions have various effectiveness in SR, and in this case, we find that student-t prior shows stronger performance than the well-known Gaussian prior. In the testing phase, we adopt the learned multiple mixture prior models to map the input LR features into the appropriate subspace, and finally reconstruct the corresponding HR image in a novel mixed matching way. Experimental results indicate that the proposed approach is both quantitatively and qualitatively superior to some state-of-the-art SR methods. [ABSTRACT FROM AUTHOR]

With the large amount of high-spatial resolution images now available, scene classification aimed at obtaining high-level semantic concepts has drawn great attention. The convolutional neural networks (CNNs), which are typical deep learning methods, have widely been studied to automatically learn features for the images for scene classification. However, scene classification based on CNNs is still difficult due to the scale variation of the objects in remote sensing imagery. In this paper, a multiscale CNN (MCNN) framework is proposed to solve the problem. In MCNN, a network structure containing dual branches of a fixed-scale net (F-net) and a varied-scale net (V-net) is constructed and the parameters are shared by the F-net and V-net. The images and their rescaled images are fed into the F-net and V-net, respectively, allowing us to simultaneously train the shared network weights on multiscale images. Furthermore, to ensure that the features extracted from MCNN are scale invariant, a similarity measure layer is added to MCNN, which forces the two feature vectors extracted from the image and its corresponding rescaled image to be as close as possible in the training phase. To demonstrate the effectiveness of the proposed method, we compared the results obtained using three widely used remote sensing data sets: the UC Merced data set, the aerial image data set, and the google data set of SIRI-WHU. The results confirm that the proposed method performs significantly better than the other state-of-the-art scene classification methods. [ABSTRACT FROM AUTHOR]

Emotion recognition is challenging due to the emotional gap between emotions and audio–visual features. Motivated by the powerful feature learning ability of deep neural networks, this paper proposes to bridge the emotional gap by using a hybrid deep model, which first produces audio–visual segment features with Convolutional Neural Networks (CNNs) and 3D-CNN, then fuses audio–visual segment features in a Deep Belief Networks (DBNs). The proposed method is trained in two stages. First, CNN and 3D-CNN models pre-trained on corresponding large-scale image and video classification tasks are fine-tuned on emotion recognition tasks to learn audio and visual segment features, respectively. Second, the outputs of CNN and 3D-CNN models are combined into a fusion network built with a DBN model. The fusion network is trained to jointly learn a discriminative audio–visual segment feature representation. After average-pooling segment features learned by DBN to form a fixed-length global video feature, a linear Support Vector Machine is used for video emotion classification. Experimental results on three public audio–visual emotional databases, including the acted RML database, the acted eNTERFACE05 database, and the spontaneous BAUM-1s database, demonstrate the promising performance of the proposed method. To the best of our knowledge, this is an early work fusing audio and visual cues with CNN, 3D-CNN, and DBN for audio–visual emotion recognition. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of recognizing human actions from depth videos. We propose a depth-based local descriptor and affine subspace coding representation with locality-constrained affine subspace coding (LASC) for 3D action recognition. First, each depth video sequence is divided into a set of subsequences (i.e., multi-scale sub-actions) based on the normalized motion energy vector. Next, depth motion map-based gradient local auto-correlation features are employed to capture the shape information and motion cues of each sub-action. In order to obtain discriminative and compact representation, we extract the local high-order information of the depth video using LASC. Through experiments, we show that the use of LASC exhibits better performance compared with existing methods such as locality-constrained linear coding. We compared LASC with the state-of-the-art methods based on similar principle, using features extracted from a single modality, on four datasets, and with those using multiple features or nonlinear recognition machines. The results on four datasets clearly show the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Classification of plants based on a multi-organ approach is very challenging. Although additional data provide more information that might help to disambiguate between species, the variability in shape and appearance in plant organs also raises the degree of complexity of the problem. Despite promising solutions built using deep learning enable representative features to be learned for plant images, the existing approaches focus mainly on generic features for species classification, disregarding the features representing plant organs. In fact, plants are complex living organisms sustained by a number of organ systems. In our approach, we introduce a hybrid generic-organ convolutional neural network (HGO-CNN), which takes into account both organ and generic information, combining them using a new feature fusion scheme for species classification. Next, instead of using a CNN-based method to operate on one image with a single organ, we extend our approach. We propose a new framework for plant structural learning using the recurrent neural network-based method. This novel approach supports classification based on a varying number of plant views, capturing one or more organs of a plant, by optimizing the contextual dependencies between them. We also present the qualitative results of our proposed models based on feature visualization techniques and show that the outcomes of visualizations depict our hypothesis and expectation. Finally, we show that by leveraging and combining the aforementioned techniques, our best network outperforms the state of the art on the PlantClef2015 benchmark. The source code and models are available at https://github.com/cs-chan/Deep-Plant. [ABSTRACT FROM AUTHOR]

Metric learning has attracted wide attention in face and kinship verification, and a number of such algorithms have been presented over the past few years. However, most existing metric learning methods learn only one Mahalanobis distance metric from a single feature representation for each face image and cannot make use of multiple feature representations directly. In many face-related tasks, we can easily extract multiple features for a face image to extract more complementary information, and it is desirable to learn distance metrics from these multiple features, so that more discriminative information can be exploited than those learned from individual features. To achieve this, we present a large-margin multi-metric learning (LM3L) method for face and kinship verification, which jointly learns multiple global distance metrics under which the correlations of different feature representations of each sample are maximized, and the distance of each positive pair is less than a low threshold and that of each negative pair is greater than a high threshold. To better exploit the local structures of face images, we also propose a local metric learning and local LM3Lmethods to learn a set of local metrics. Experimental results on three face data sets show that the proposed methods achieve very competitive results compared with the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Individuating and locating repetitive patterns in still images is a fundamental task in image processing, typically achieved by means of correlation strategies. In this paper, we provide a solid solution to this task using a differential geometry approach, operating on Lie algebra, and exploiting a mixture of templates. The proposed method asks the user to locate a few instances of the target patterns (seeds) that become visual templates used to explore the image. We propose an iterative algorithm to locate patches similar to the seeds working in three steps: first, clustering the detected patches to generate templates of different classes, then looking for the affine transformations, living on a Lie algebra that best links the templates and the detected patches, and finally detecting new patches with a convolutional strategy. The process ends when no new patches are found. We will show how our method is able to process heterogeneous unstructured images with multiple visual motifs and extremely crowded scenarios with high precision and recall, outperforming all the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In this paper, we propose a nonlinear structural hashing approach to learn compact binary codes for scalable video search. Unlike most existing video hashing methods which consider image frames within a video separately for binary code learning, we develop a multi-layer neural network to learn compact and discriminative binary codes by exploiting both the structural information between different frames within a video and the nonlinear relationship between video samples. To be specific, we learn these binary codes under two different constraints at the output of our network: 1) the distance between the learned binary codes for frames within the same scene is minimized and 2) the distance between the learned binary matrices for a video pair with the same label is less than a threshold and that for a video pair with different labels is larger than a threshold. To better measure the structural information of the scenes from videos, we employ a subspace clustering method to cluster frames into different scenes. Moreover, we design multiple hierarchical nonlinear transformations to preserve the nonlinear relationship between videos. Experimental results on three video data sets show that our method outperforms state-of-the-art hashing approaches on the scalable video search task. [ABSTRACT FROM AUTHOR]

Estimating nutrient content in plants is a crucial task in the application of precision farming. This work will be more challenging if it is conducted nondestructively based on plant images captured in the field due to the variation of lighting conditions. This paper proposes a computational intelligence image processing to analyze nitrogen status in wheat plants. We developed an ensemble of deep learning multilayer perceptron-using committee machines for color normalization and image segmentation. This paper also focuses on building a genetic-algorithm-based global optimization to fine tune the color normalization and nitrogen estimation results. We discovered that the proposed method can successfully normalize plant images by reducing color variabilities compared to other color normalization techniques. Furthermore, this algorithm is able to enhance the nitrogen estimation results compared to other non-global optimization methods as well as the most renowned SPAD meter based nitrogen measurement. [ABSTRACT FROM AUTHOR]

The performance of scene classification relies heavily on the spatial and structural features that are extracted from high spatial resolution remote-sensing images. Existing approaches, however, are limited in adequately exploiting latent relationships between scene images. Aiming to decrease the distances between intraclass images and increase the distances between interclass images, we propose a latent relationship learning framework that integrates an adaptive graph with the constraints of the feature space and label propagation for high-resolution aerial image classification. To describe the latent relationships among scene images in the framework, we construct an adaptive graph that is embedded into the constrained joint space for features and labels. To remove redundant information and improve the computational efficiency, subspace learning is introduced to assist in the latent relationship learning. To address out-of-sample data, linear regression is adopted to project the semisupervised classification results onto a linear classifier. Learning efficiency is improved by minimizing the objective function via the linearized alternating direction method with an adaptive penalty. We test our method on three widely used aerial scene image data sets. The experimental results demonstrate the superior performance of our method over the state-of-the-art algorithms in aerial scene image classification. [ABSTRACT FROM AUTHOR]

In remote sensing image classification, active learning aims to learn a good classifier as best as possible by choosing the most valuable (informative and representative) training samples. Multiview is a concept that regards analyzing the same object from multiple different views. Generally, these views show diversity and complementarity of features. In this paper, we propose a new multiview active learning (MVAL) framework for hyperspectral image classification. First, we generate multiple views by extracting different attribute components from the same image data. Specifically, we adopt the multiple morphological component analysis to decompose the original image into multiple pairs of attribute components, including content, coarseness, contrast, and directionality, and the smooth component from each pair is chosen as one single view. Second, we construct two multiview intensity-based query strategies for active learning. On the one hand, we exploit the intensity differences of multiple views along with the samples' uncertainty to choose the most informative candidates. On the other hand, we consider the clustering distribution of all unlabeled samples, and query the most representative candidates in addition to the highly informative ones. Our experiments are performed on four benchmark hyperspectral image data sets. The obtained results show that the proposed MVAL framework can lead to better classification performance than the traditional, single-view active learning schemes. In addition, compared with the conventional disagree-based MVAL scheme, the proposed query selection strategies show competitive classification accuracy. [ABSTRACT FROM AUTHOR]

In this paper we present Latent-Class Hough Forests, a method for object detection and 6 DoF pose estimation in heavily cluttered and occluded scenarios. We adapt a state of the art template matching feature into a scale-invariant patch descriptor and integrate it into a regression forest using a novel template-based split function. We train with positive samples only and we treat class distributions at the leaf nodes as latent variables. During testing we infer by iteratively updating these distributions, providing accurate estimation of background clutter and foreground occlusions and, thus, better detection rate. Furthermore, as a by-product, our Latent-Class Hough Forests can provide accurate occlusion aware segmentation masks, even in the multi-instance scenario. In addition to an existing public dataset, which contains only single-instance sequences with large amounts of clutter, we have collected two, more challenging, datasets for multiple-instance detection containing heavy 2D and 3D clutter as well as foreground occlusions. We provide extensive experiments on the various parameters of the framework such as patch size, number of trees and number of iterations to infer class distributions at test time. We also evaluate the Latent-Class Hough Forests on all datasets where we outperform state of the art methods. [ABSTRACT FROM AUTHOR]

Kernel principal component analysis (KPCA) is a powerful and widely applied nonlinear feature extraction technique. However, as originally proposed, KPCA may be cumbersome or infeasible in large-scale datasets, which motivated the development of low-complexity iterative extraction algorithms, mainly aiming image processing applications. Recently, some online KPCA extraction algorithms were proposed, but most of them suffer from low-convergence speed. This paper proposes a new algorithm based on fixed-point iterative equations for KPCA extraction, expanding kernel components using a compact dictionary, dynamically built from data, according to a user-defined accuracy parameter. The algorithm relies on simple equations, can track nonstationary environments, and requires reduced storage, enabling its use in real-time applications operating in low-cost embedded hardware. Results involving open-access image datasets show improved accuracy and convergence speed, as well as permitted effective improvements in practical image applications, as compared to state-of-art online KPCA techniques. [ABSTRACT FROM PUBLISHER]

Biometric applications have been used globally in everyday life. However, conventional biometrics is created and optimized for high-security scenarios. Being used in daily life by ordinary untrained people is a new challenge. Facing this challenge, designing a biometric system with prior constraints of ergonomics, we propose ergonomic biometrics design model, which attains the physiological factors, the psychological factors, and the conventional security characteristics. With this model, a novel hand-based biometric system, door knob hand recognition system (DKHRS), is proposed. DKHRS has the identical appearance of a conventional door knob, which is an optimum solution in both physiological factors and psychological factors. In this system, a hand image is captured by door knob imaging scheme, which is a tailored omnivision imaging structure and is optimized for this predetermined door knob appearance. Then features are extracted by local Gabor binary pattern histogram sequence method and classified by projective dictionary pair learning. In the experiment on a large data set including 12 000 images from 200 people, the proposed system achieves competitive recognition performance comparing with conventional biometrics like face and fingerprint recognition systems, with an equal error rate of 0.091%. This paper shows that a biometric system could be built with a reliable recognition performance under the ergonomic constraints. [ABSTRACT FROM AUTHOR]

This paper aims for event recognition when video examples are scarce or even completely absent. The key in such a challenging setting is a semantic video representation. Rather than building the representation from individual attribute detectors and their annotations, we propose to learn the entire representation from freely available web videos and their descriptions using an embedding between video features and term vectors. In our proposed embedding, which we call Video2vec, the correlations between the words are utilized to learn a more effective representation by optimizing a joint objective balancing descriptiveness and predictability. We show how learning the Video2vec embedding using a multimodal predictability loss, including appearance, motion and audio features, results in a better predictable representation. We also propose an event specific variant of Video2vec to learn a more accurate representation for the words, which are indicative of the event, by introducing a term sensitive descriptiveness loss. Our experiments on three challenging collections of web videos from the NIST TRECVID Multimedia Event Detection and Columbia Consumer Videos datasets demonstrate: i) the advantages of Video2vec over representations using attributes or alternative embeddings, ii) the benefit of fusing video modalities by an embedding over common strategies, iii) the complementarity of term sensitive descriptiveness and multimodal predictability for event recognition. By its ability to improve predictability of present day audio-visual video features, while at the same time maximizing their semantic descriptiveness, Video2vec leads to state-of-the-art accuracy for both few- and zero-example recognition of events in video. [ABSTRACT FROM PUBLISHER]