YOLSO: You Only Look Small Object

Small object detection is challenging and far from satisfactory. Most general object detectors suffer from two critical issues with small objects: (1) Feature extractor based on classification network cannot express the characteristics of small objects reasonably due to insufficient appearance information of targets and a large amount of background interference around them. (2) The detector requires a much higher location accuracy for small objects than for general objects. This paper proposes an effective and efficient small object detector YOLSO to address the above problems. For feature representation, we analyze the drawbacks in previous backbones and present a Half-Space Shortcut(HSSC) module to build a background-aware backbone. Furthermore, a coarse-to-fine Feature Pyramid Enhancement(FPE) module is introduced for layer-wise aggregation at a granular level to enhance the semantic discriminability. For loss function, we propose an exponential L1 loss to promote the convergence of regression, and a focal IOU loss to focus on prime samples with high classification confidence and high IOU. Both of them significantly improves the location accuracy of small objects. The proposed YOLSO sets state-of-the-art results on two typical small object datasets, MOCOD and VeDAI, at a speed of over 200 FPS. In the meantime, it also outperforms the baseline YOLOv3 by a wide margin on the common COCO dataset.     

Finite-time adaptive fuzzy control for a class of output constrained nonlinear systems with dead-zone

The article investigates the finite-time adaptive fuzzy control for a class of nonlinear systems with output constraint and input dead-zone. First, by skillfully combining the barrier Lyapunov function, backstepping design method, and finite-time control theory, a novel adaptive state-feedback tracking controller is constructed, and the output constraint of the nonlinear system is not violated. Second, the fuzzy logic system is used to approximate unknown function in the nonlinear system. Third, the finite-time command filter is introduced to avoid the problem of “complexity explosion” caused by repeated differentiations of the virtual control signal in conventional backstepping control schemes. Meanwhile, a new saturation function is added in the compensating signal for filter error to improve control accuracy. Finally, based on Lyapunov stability analysis, all the signals of the closed-loop are proved to be semi-globally uniformly ultimately bounded, and the tracking error converges to a small neighborhood region of the origin in a finite time. A simulation example is presented to demonstrate the effectiveness for the proposed control scheme.     

Identification and U-control of a state-space system with time-delay

This article presents a state-space model with time-delay to map the relationship between known input-output data for discrete systems. For the given input-output data, a model identification algorithm combining parameter estimation and state estimation is proposed in line with the causality constraints. Consequently, this article proposes a least squares parameter estimation algorithm, and analyzes its convergence for the studied systems to prove that the parameter estimation errors converge to zero under the persistent excitation conditions. In control system design, the U-model based control is introduced to provide a unilateral platform to improve the design efficiency and generality. A simulation portfolio from modeling to control is provided with computational experiments to validate the derived results.     

High quality 3D reconstruction based on fusion of polarization imaging and binocular stereo vision

Polarization imaging can retrieve inaccurate objects’ 3D shapes with fine textures, whereas coarse but accurate depths can be provided by binocular stereo vision. To take full advantage of these two complementary techniques, we investigate a novel 3D reconstruction method based on the fusion of polarization imaging and binocular stereo vision for high quality 3D reconstruction. We first generate the polarization surface by correcting the azimuth angle errors on the basis of registered binocular depth, to solve the azimuthal ambiguity in the polarization imaging. Then we propose a joint 3D reconstruction model for depth fusion, including a data fitting term and a robust low-rank matrix factorization constraint. The former is to transfer textures from the polarization surface to the fused depth by assuming their relationship linear, whereas the latter is to utilize the low-frequency part of binocular depth to improve the accuracy of the fused depth considering the influences of missing-entries and outliers. To solve the optimization problem in the proposed model, we adopt an efficient solution based on the alternating direction method of multipliers. Extensive experiments have been conducted to demonstrate the efficiency of the proposed method in comparison with state-of-the-art methods and to exhibit its wide application prospects in 3D reconstruction.     

基于误差估值累加开环校正的诱导式欺骗检测方法

诱导式卫星欺骗干扰可诱导航空器逐渐偏离预定航迹，难以被发现，因此及时有效地检测干扰是飞行安全的保障。在现有紧组合导航体制基础上，设计了一种基于误差估值累加开环校正的紧组合导航结构，并证明了其性能与传统闭环校正紧组合导航性能等效。在此结构中，将紧组合导航系统与自适应序贯概率比检测方法结合，提出了一种基于误差估值累加开环校正的诱导式欺骗检测方法，融合紧组合导航信息与其他不受欺骗影响的导航信息，构建欺骗检测统计量进行诱导式欺骗检测。仿真结果表明，开环校正结构可避免随时间累加的惯性导航系统误差所导致的组合导航滤波器发散问题，同时欺骗检测方法可进一步提高算法对“最坏”情形下微小诱导式欺骗的检测效果。     

Data fusion and transfer learning empowered granular trust evaluation for Internet of Things

In the Internet of Things (IoT), a huge amount of valuable data is generated by various IoT applications. As the IoT technologies become more complex, the attack methods are more diversified and can cause serious damages. Thus, establishing a secure IoT network based on user trust evaluation to defend against security threats and ensure the reliability of data source of collected data have become urgent issues, in this paper, a Data Fusion and transfer learning empowered granular Trust Evaluation mechanism (DFTE) is proposed to address the above challenges. Specifically, to meet the granularity demands of trust evaluation, time–space empowered fine/coarse grained trust evaluation models are built utilizing deep transfer learning algorithms based on data fusion. Moreover, to prevent privacy leakage and task sabotage, a dynamic reward and punishment mechanism is developed to encourage honest users by dynamically adjusting the scale of reward or punishment and accurately evaluating users’ trusts. The extensive experiments show that: (i) the proposed DFTE achieves high accuracy of trust evaluation under different granular demands through efficient data fusion; (ii) DFTE performs excellently in participation rate and data reliability.     

Benchmark experiment of large-angle scattering reaction cross section of iron at 14 MeV using two shadow bars – Comparison of experimental results with ENDF/B-VIII –

ABSTRACT

It is important to perform neutron transport simulations with accurate nuclear data in the neutronics design of a fusion reactor. However, absolute values of large-angle scattering cross sections vary among nuclear data libraries even for well-examined nuclide of iron. Benchmark experiments focusing on large-angle scattering cross sections were thus performed to confirm the correctness of nuclear data libraries. The series benchmark experiments were performed at a DT neutron source facility, OKTAVIAN of Osaka University, Japan, by the unique experimental system established by the authors’ group, which can extract only the contribution of large-angle scattering reactions. This system consists of two shadow bars, target plate (iron), and neutron detector (niobium). Two types of shadow bars were used and four irradiations were conducted for one experiment, so that contribution of room-return neutrons was effectively removed and only large-angle scattering neutrons were extracted from the measured four Nb reaction rates. The obtained experimental results were compared with calculations for five nuclear data libraries including JENDL-4.0, JEFF.-3.3, FENDL-3.1, ENDF/B- VII, and recently released ENDF/B-VIII. It was found from the comparison that ENDF/B-VIII showed the best result, though ENDF/B-VII showed overestimation and others are in large underestimation at 14 MeV.     

Feature selection methods for root-cause analysis among top-level product attributes

Manufacturing companies not only strive to deliver flawless products but also monitor product failures in the field to identify potential quality issues. When product failures occur, quality engineers must identify the root cause to improve any affected product and process. This root-cause analysis can be supported by feature selection methods that identify relevant product attributes, such as manufacturing dates with an increased number of product failures. In this paper, we present different methods for feature selection and evaluate their ability to identify relevant product attributes in a root-cause analysis. First, we compile a list of feature selection methods. Then, we summarize the properties of product attributes in warranty case data and discuss these properties regarding the challenges they pose for machine learning algorithms. Next, we simulate datasets of warranty cases, which emulate these product properties. Finally, we compare the feature selection methods based on these simulated datasets. In the end, the univariate filter information gain is determined to be a suitable method for a wide range of applications. The comparison based on simulated data provides a more general result than other publications, which only focus on a single use case. Due to the generic nature of the simulated datasets, the results can be applied to various root-cause analysis processes in different quality management applications and provide a guideline for readers who wish to explore machine learning methods for their analysis of quality data.     

自适应融合层级特征的混合退化图像复原算法

多种退化类型混合的图像比单一类型的退化图像降质更严重,很难建立精确模型对其复原,研究端到端的神经网络算法是复原的关键.现有的基于操作选择注意力网络的算法(operation-wiseattentionnetwork,OWAN)虽然有一定的性能提升,但是其网络过于复杂,运行较慢,复原图像缺乏高频细节,整体效果也有提升的空间.针对这些问题,提出一种基于层级特征融合的自适应复原算法.该算法直接融合不同感受野分支的特征,增强复原图像的结构;用注意力机制对不同层级的特征进行动态融合,增加模型的自适应性,降低了模型冗余;另外,结合L1损失和感知损失,增强了复原图像的视觉感知效果.在DIV2K,BSD500等数据集上的实验结果表明,该算法无论是在峰值信噪比和结构相似性上的定量分析,还是在主观视觉质量方面,均优于OWAN算法,充分证明了该算法的有效性.     

A novel fusion paradigm for multi-channel image denoising

Multi-channel and single-channel image denoising are on two important development fronts. Integrating multi-channel and single-channel image denoisers for further improvement is a valuable research direction. A natural assumption is that using more useful information is helpful to the output results. In this paper, a novel multi-channel and single-channel fusion paradigm (MSF) is proposed. The proposed MSF works by fusing the estimates of a multi-channel image denoiser and a single-channel image denoiser. The performance of recent multi-channel image denoising methods involved in the proposed MSF can be further improved at low additional time-consuming cost. Specifically, the validity principle of the proposed MSF is that the fused single-channel image denoiser can produce auxiliary estimate for the involved multi-channel image denoiser in a designed underdetermined transform domain. Based on the underdetermined transformation, we create a corresponding orthogonal transformation for fusion and better restore the multi-channel images. The quantitative and visual comparison results demonstrate that the proposed MSF can be effectively applied to several state-of-the-art multi-channel image denoising methods.