Fine-grained visual categorization (FGVC) has drawn increasing attention as an emerging research field in recent years. In contrast to generic-domain visual recognition, FGVC is characterized by high intra-class and subtle inter-class variations. To distinguish conceptually and visually similar categories, highly discriminative visual features must be extracted. Moreover, FGVC has highly specialized and task-specific nature. It is not always easy to obtain a sufficiently large-scale training dataset. Therefore, the key to success in practical FGVC systems is to efficiently exploit discriminative features from a limited number of training examples. In this paper, we propose an efficient two-step dimensionality compression method to derive compact middle-level part-based features. To do this, we compare both space-first and feature-first convolution schemes and investigate their effectiveness. Our approach is based on simple linear algebra and analytic solutions, and is highly scalable compared with the current one-vs-one or one-vs-all approach, making it possible to quickly train middle-level features from a number of pairwise part regions. We experimentally show the effectiveness of our method using the standard Caltech-Birds and Stanford-Cars datasets.

A non-linear extension of generalized hyperplane approximation (GHA) method is introduced in this letter. Although GHA achieved a high-confidence result in motion parameter estimation by utilizing the supervised learning scheme in histogram of oriented gradient (HOG) feature space, it still has unstable convergence range because it approximates the non-linear function of regression from the feature space to the motion parameter space as a linear plane. To extend GHA into a non-linear regression for larger convergence range, we derive theoretical equations and verify this extension's effectiveness and efficiency over GHA by experimental results.

Key frame based video summarization has emerged as an important task for efficient video data management. This paper proposes a novel technique for key frame extraction based on chaos theory and color information. By applying chaos theory, a large content change between frames becomes more chaos-like and results in a more complex fractal trajectory in phase space. By exploiting the fractality measured in the phase space between frames, it is possible to evaluate inter-frame content changes invariant to effects of fades and illumination change. In addition to this measure, the color histogram-based measure is also used to complement the chaos-based measure which is sensitive to changes of camera /object motion. By comparing the last key frame with the current frame based on the proposed frame difference measure combining these two complementary measures, the key frames are robustly selected even under presence of video fades, changes of illumination, and camera/object motion. The experimental results demonstrate its effectiveness with significant improvement over the conventional method.

Based on the completeness of the real-valued discrete Gabor transform, a new biorthogonal relationship between analysis window and synthesis window is derived and a fast algorithm for computing the analysis window is presented for any given synthesis window. The new biorthogonal relationship can be expressed as a linear equation set, which can be separated into a certain number of independent sub-equation sets, where each of them can be fast and independently solved by using convolution operations and FFT to obtain the analysis window for any given synthesis window. Computational complexity analysis and comparison indicate that the proposed algorithm can save a considerable amount of computation and is more efficient than the existing algorithms.

In recent years, optical music recognition (OMR) has been extensively developed, particularly for use with mobile devices that require fast processing to recognize and play live the notes in images captured from sheet music. However, most techniques that have been developed thus far have focused on playing back instrumental music and have ignored the importance of lyric extraction, which is time consuming and affects the accuracy of the OMR tools. The text of the lyrics adds complexity to the page layout, particularly when lyrics touch or overlap musical symbols, in which case it is very difficult to separate them from each other. In addition, the distortion that appears in captured musical images makes the lyric lines curved or skewed, making the lyric extraction problem more complicated. This paper proposes a new approach in which lyrics are detected and extracted quickly and effectively. First, in order to resolve the distortion problem, the image is undistorted by a method using information of stave lines and bar lines. Then, through the use of a frequency count method and heuristic rules based on projection, the lyric areas are extracted, the cases where symbols touch the lyrics are resolved, and most of the information from the musical notation is kept even when the lyrics and music notes are overlapping. Our algorithm demonstrated a short processing time and remarkable accuracy on two test datasets of images of printed Korean musical scores: the first set included three hundred scanned musical images; the second set had two hundred musical images that were captured by a digital camera.

This article presents a novel, approximate method that makes it possible to analyse multi-service switching networks. The method belongs to the group of the so-called effective availability methods and is characterized by very high accuracy for single-service and multi-service switching networks alike. The operation of the proposed method is presented with an example of a number of three-stage switching networks with different ways of the execution of inter-stage connections. A comparison of analytical and simulation results confirms high accuracy of the proposed method that is independent of the structure of a switching network.

Because the perceptual compressive sensing framework can achieve a much better performance than the legacy compressive sensing framework, it is very promising for the compressive sensing based image compression system. In this paper, we propose an innovative adaptive perceptual block compressive sensing scheme. Firstly, a new block-based statistical metric which can more appropriately measure each block's sparsity and perceptual sensibility is devised. Then, the approximated theoretical minimum measurement number for each block is derived from the new block-based metric and used as weight for adaptive measurements allocation. The obtained experimental results show that our scheme can significantly enhance both objective and subjective performance of a perceptual compressive sensing framework.

This study focuses on modeling the storytelling performance of the participants in a group conversation. Storytelling performance is one of the fundamental communication techniques for providing information and entertainment effectively to a listener. We present a multimodal analysis of the storytelling performance in a group conversation, as evaluated by external observers. A new multimodal data corpus is collected through this group storytelling task, which includes the participants' performance scores. We extract multimodal (verbal and nonverbal) features regarding storytellers and listeners from a manual description of spoken dialog and from various nonverbal patterns, including each participant's speaking turn, utterance prosody, head gesture, hand gesture, and head direction. We also extract multimodal co-occurrence features, such as head gestures, and interaction features, such as storyteller utterance overlapped with listener's backchannel. In the experiment, we modeled the relationship between the performance indices and the multimodal features using machine-learning techniques. Experimental results show that the highest accuracy (R2) is 0.299 for the total storytelling performance (sum of indices scores) obtained with a combination of verbal and nonverbal features in a regression task.

For a service-oriented architecture-based system, the problem of synthesizing a concrete model (i.e., a behavioral model) for each peer configuring the system from an abstract specification — which is referred to as choreography — is known as the choreography realization problem. In this paper, we consider the condition for the behavioral model when choreography is given by an acyclic relation. A new notion called re-constructible decomposition of acyclic relations is introduced, and a necessary and sufficient condition for a decomposed relation to be re-constructible is shown. The condition provides lower and upper bounds of the acyclic relation for the behavioral model. Thus, the degree of freedom for behavioral models increases; developing algorithms for synthesizing an intelligible model for users becomes possible. It is also expected that the condition is applied to the case where choreography is given by a set of acyclic relations.

Connection Service Providers (CSP) are wishing to increase their Return on Investment (ROI) by utilizing the data assets generated by tracking subscriber behaviors. This results in the ability to apply personalized policies, monitor and control the service traffic to subscribers and gain more revenue through the usage of subscriber data with ad networks. In this paper, a system is proposed to monitor and analyze the Internet access of the subscribers of a regional SP in order to classify the subscribers into interest categories from the Interactive Advertising Bureau (IAB) categories. The study employs the categorization engine to build category vectors for all individuals using Internet services through the subscription. The proposal makes it easy to detect changes in the interests of individuals/subscribers over time.

This paper presents analysis and simple design guideline for ThruChip Interface (TCI) as located by LC-VCO which is used in high-speed SoC. The electromagnetic interference (EMI) from TCI channels to LC-VCO is analyzed and evaluated. The accuracy of the analysis and design guidelines is verified through the test-chip verification.

Residents living in a nursing home usually have established medical histories in multiple sources, and most previous medicine management systems have only focused on the integration of prescriptions and the identification of repeated drug uses. Therefore, a comprehensive medicine management system is proposed to integrate medical information from different sources. The proposed system not only detects inappropriate drugs automatically but also allows users to input such information for any non-prescription medicines that the residents take. Every participant can fully track the residents' latest medicine use online and in real time. Pharmacists are able to issue requests for suggestions on medicine use, and residents can also have a comprehensive understanding of their medicine use. The proposed scheme has been practically implemented in a nursing home in Taiwan. The evaluation results show that the average time to detect an inappropriate drug use and complete a medicine record is reduced. With automatic and precise comparisons, the repeated drugs and drug side effects are identified effectively such that the amount of medicine cost spent on the residents is also reduced. Consequently, the proactive feedback, real-time tracking, and interactive consulting mechanisms bind all parties together to realize a comprehensive medicine management system.

User-based and item-based collaborative filtering (CF) are two of the most important and popular techniques in recommender systems. Although they are widely used, there are still some limitations, such as not being well adapted to the sparsity of data sets, failure to consider the hierarchical structure of the items, and changes in users' interests when calculating the similarity of items. To overcome these shortcomings, we propose an evolutionary approach based on hierarchical structure for dynamic recommendation system named Hierarchical Temporal Collaborative Filtering (HTCF). The main contribution of the paper is displayed in the following two aspects. One is the exploration of hierarchical structure between items to improve similarity, and the other is the improvement of the prediction accuracy by utilizing a time weight function. A unique feature of our method is that it selects neighbors mainly based on hierarchical structure between items, which is more reliable than co-rated items utilized in traditional CF. To the best of our knowledge, there is little previous work on researching CF algorithm by combining object implicit or latent object-structure relations. The experimental results show that our method outperforms several current recommendation algorithms on recommendation accuracy (in terms of MAE).

In this paper, we investigate a preference-aware multicast mechanism in active array aided LTE (Long Term Evolution) networks. An active antenna system can direct vertical beams in different horizontal and vertical directions, so the amount of energy delivered is more concentrated on the target users. The active array provides each multicast group with an individual beam with specific downtilt delivering shared video to all users in the group. For the multicast system, the objective of our proposed resource allocation scheme is to maximize the total throughput, subject to the constraints of power, subcarrier and antenna downtilt, as well as horizontal angles and the vertical half power bandwidth. To solve the problem, individual beams are steered for multicast groups. Furthermore, a novel subcarrier assignment scheme is proposed to enhance the spectrum resource utilization, and the optimal power allocation is obtained by virtue of Lagrangian method. Simulation results demonstrate the throughput and the spectral efficiency enhancement of our proposed scheme over other conditional schemes.

We describe a calculation tool and modeling methods to find self and mutual inductance and current distribution in superconductive multilayer circuit layouts. Accuracy of the numerical solver is discussed and compared with experimental measurements. Effects of modeling parameter selection on calculation results are shown, and we make conclusions on the selection of modeling parameters for fast but sufficiently accurate calculations when calibration methods are used. Circuit theory for the calculation of branch impedances from the output of the numerical solver is discussed, and compensation for solution difficulties is shown through example. We elaborate on the construction of extraction models for superconductive integrated circuits, with and without resistive branches. We also propose a method to calculate current distribution in a multilayer circuit with multiple bias current feed points. Finally, detailed examples are shown where the effects of stacked vias, bias pillars, coupling, ground connection stacks and ground return currents in circuit layouts for the AIST advanced process (ADP2) and standard process (STP2) are analyzed. We show that multilayer inductance and current distribution extraction in such circuits provides much more information than merely branch inductance, and can be used to improve layouts; for example through reduced coupling between conductors.

Designing a backbone IP network, especially to support both unicast and multicast traffic under delay constraints, is a difficult problem. Real network design must consider cost, performance and reliability. Therefore, a simulator can help a network designer to test the functionality of the network before the implementation. This paper proposes a heuristic design algorithm called D-MENTOR, and the algorithm was developed by programming based on Mesh Network Topological Optimization and Routing Version 2 (MENTOR-II) to integrate as a new module of DElite tool. The simulation results show that, in almost all test cases, the proposed algorithm yields lower installation cost.

Some diffusive and recurrence properties of Lorentz Lattice Gas Cellular Automata (LLGCA) have been expensively studied in terms of the densities of some of the left/right static/flipping mirrors/rotators. In this paper, for any combination S of these well known scatters, we study the computational complexity of the following problem which we call PERIODICITY on the S-model: given a finite configuration that distributes only those scatters in S, whether a particle visits the starting position periodically or not. Previously, the flipping mirror model and the occupied flipping rotator model have been shown unbounded, i.e. the process is always diffusive [17]. On the other hand, PERIODICITY is shown PSPACE-complete in the unoccupied flipping rotator model [21]. In this paper, we show that PERIODICITY is PSPACE-compete in any S-model that is neither occupied, unbounded, nor static. Particularly, we prove that PERIODICITY in any unoccupied and bounded model containing flipping mirror is PSPACE-complete.

Considering the obvious bias of the traditional interpolation method, a novel time delay estimation (TDE) interpolation method with sub-sample accuracy is presented in this paper. The proposed method uses a generalized extended approximation method to obtain the objection function. Then the optimized interpolation curve is generated by Second-order Cone programming (SOCP). Finally the optimal TDE can be obtained by interpolation curve. The delay estimate of proposed method is not forced to lie on discrete samples and the sample points need not to be on the interpolation curve. In the condition of the acceptable computation complexity, computer simulation results clearly indicate that the proposed method is less biased and outperforms the other interpolation algorithms in terms of estimation accuracy.

Distributed compressive video sensing (DCVS) is an emerging low-complexity video coding framework which integrates the merits of distributed video coding (DVC) and compressive sensing (CS). In this paper, we propose a novel rate-distortion optimized DCVS codec, which takes advantage of a rate-distortion optimization (RDO) model based on the estimated correlation noise (CN) between a non-key frame and its side information (SI) to determine the optimal measurements allocation for the non-key frame. Because the actual CN can be more accurately recovered by our DCVS codec, it leads to more faithful reconstruction of the non-key frames by adding the recovered CN to the SI. The experimental results reveal that our DCVS codec significantly outperforms the legacy DCVS codecs in terms of both objective and subjective performance.

A generalized covariance matrix taper (GCMT) model is proposed to enhance the performance of knowledge-aided space-time adaptive processing (KA-STAP) under sea clutter environments. In KA-STAP, improving the accuracy degree of the a priori clutter covariance matrix is a fundamental issue. As a crucial component in the a priori clutter covariance matrix, the taper matrix is employed to describe the internal clutter motion (ICM) or other subspace leakage effects, and commonly constructed by the classical covariance matrix taper (CMT) model. This work extents the CMT model into a generalized CMT (GCMT) model with a greater degree of freedom. Comparing it with the CMT model, the proposed GCMT model is more suitable for sea clutter background applications for its improved flexibility. Simulation results illustrate the efficiency of the GCMT model under different sea clutter environments.